Board logo

Composite chassis
ceebmoj - 14/10/09 at 12:50 PM

This morning I saw a GTM on the way to work witch got me thinking about Composite chassis. Does any body have a link to the manufacture of a Composite chassis GTM or other. There was thread in the three wheal section a while back but I cant find that now.


djtom - 14/10/09 at 01:15 PM

Westfield made a composite chassis car a few years ago. FW400 I think it was called. 400kg but OMG expensive.


MakeEverything - 14/10/09 at 01:35 PM

Id rather like the idea of a monocoque carbon fibre chassis for my kit..... if i had the money!


ceebmoj - 14/10/09 at 02:33 PM

quote:
Originally posted by MakeEverything
Id rather like the idea of a monocoque carbon fibre chassis for my kit..... if i had the money!


I guess that s what got me thinking the GTM is GRP so I assume it responsible to manufacture I will keep having a look for the GRP over fome trike chassy that is out there


Triggerhappy - 14/10/09 at 03:11 PM

Send your drawings to these guy´s
http://www.atrgroup.it/

They have some experience.


iank - 14/10/09 at 04:45 PM

Think this is the thread you're talking about
http://www.locostbuilders.co.uk/viewthread.php?tid=94561

The link to the trimagnum grp over foam approach is here
http://www.rqriley.com/frp-foam.htm
Note there is some chassis under there, more of a big subframe really.

smart51's cabin scooter uses this method of construction, but also has a chassis. Lots of good information in his posts and photos.

[Edited on 14/10/09 by iank]


ceebmoj - 14/10/09 at 05:28 PM

quote:
Originally posted by iank
Think this is the thread you're talking about
http://www.locostbuilders.co.uk/viewthread.php?tid=94561

The link to the trimagnum grp over foam approach is here
http://www.rqriley.com/frp-foam.htm
Note there is some chassis under there, more of a big subframe really.

smart51's cabin scooter uses this method of construction, but also has a chassis. Lots of good information in his posts and photos.

[Edited on 14/10/09 by iank]


there was anothere 1 + 1 three wheeler that was all grp over foam with bonded in kevlar bulkheads

[Edited on 14/10/09 by ceebmoj]


clairetoo - 14/10/09 at 06:08 PM

quote:
Originally posted by ceebmoj

there was anothere 1 + 1 three wheeler that was all grp over foam with bonded in kevlar bulkheads

[Edited on 14/10/09 by ceebmoj]

Do you mean the TREV ? Linkydink


ceebmoj - 14/10/09 at 06:38 PM

quote:
Originally posted by clairetoo
Do you mean the TREV ? Linkydink


thanks alot that is exactly the one I was thinking about


StevieB - 14/10/09 at 06:50 PM

I've been reading about trev and have made some enquiries as to the cost for UK supply of the honeycomb grp material - no answer as yet.

What I have found though is this

Not sure if you'd get the strength, but if you gave it adequate skinning with carbon, kevlar or woven roving then a decent strength might be achievable.

I had also been considering how to vaccuum bag the a monocoque using resin infusion - not only neater but I feel it would be stronger as well because you'd have less risk of getting any air pockets trapped in there.


ceebmoj - 14/10/09 at 07:33 PM

nasty bob from the mid engen section

http://www.locostbuilders.co.uk/viewthread.php?tid=73075

has used a skined hunny come sections to pannel in parts of his chasy maby the product he used is apropreate


ceebmoj - 14/10/09 at 07:39 PM

also have you got any picture of how hard points are mounted to trev?

on old formula vauxall lotuses a hole is drilled and a aluum top hat inserted from both sides so tha you cant crush the center section.


StevieB - 14/10/09 at 08:08 PM

That's about it - ali bobbins and also in some placs a large ali spreader panel, as well as angle in the corners etc.

No details on Trev, but have a read of 'How to Build Motorcycle Engined Race Cars' and it hs some explanation, as well as Allan Staniforths 'Race and Rally Car Source Book'. The first book is OK - not much explanation of actually building a cr but a decent bit of starter info and plenty of ideas on how to do various things - it has a chapter on making a monocoque using honeycomb ali panels.

I have a distant plan to do something very similar but using the grp honeycomb/kevlar laminate along the same lines as Trev.


ceebmoj - 15/10/09 at 07:33 AM

as in this book?

http://www.amazon.co.uk/s/ref=nb_ss?url=search-alias%3Daps&field-keywords=How+to+Build+Motorcycle+Engined+Race+Car&x=0&y=0


StevieB - 15/10/09 at 12:07 PM

That's the one.

It's OK, but there again I picked up my copy for £6 (brand new) from an autojumble.

I also managed to get hold of an original, unopened copy of High Speed, Low Cost for £25 from the same guy


Theshed - 15/10/09 at 02:37 PM

My car is made out of honeycombe. A few bad pics in my folder. If I ever did it again I would make jigs and have inserts placed beteen the sheets and use laser cutting. I would also use solid bulkheads - or maybe just steel tube???

On the ther hand the locost shape might lend itself to 2 substantial side planks of honeycombe. I would go for 40mm 1/8" cell carbon/kevlar - no need to paint it either!!

I got my sheets from Aim Composites - they were good to deal with and were happy to make sheets up in Kevlar/carbon. They would have put in inserts. None of this was what you would call Lo Cost


StevieB - 15/10/09 at 04:52 PM

Right, just got my email quotation from Ayres (the company that does the stuff Trev is made from).

372 euros per sheet, which sounds quite a lot but by my figuring I reckon a single seater would be made from 3 sheets, so a monocoque single seater chassis for £1k (plus the cost of laminating in carbon/kevlar).

Probably need to price it up using cheaper materials because it'd be a bloody shame to make something that epxensive to fin a fundemental flaw in the design that needs changing (ie, a hole new chassi building).

Can you laminate ali panels with compsite materials successfully/strongly?


ceebmoj - 15/10/09 at 05:52 PM

what about the Nidaplast you posted about before it also come in a other thickness as well

[Edited on 15/10/09 by ceebmoj]


StevieB - 15/10/09 at 06:00 PM

I'm just not sure how string the nidaplast would be compared to ali honeycomb.

I might try an experiment by buying some and a small quantity of carbon, kevlar, carbon/kevlar and woven roving and see what works.


Theshed - 15/10/09 at 09:09 PM

I think that with care and preferably some heat you could get a decent bond between aluminium and Carbon/kevlar even in a diy situation. Just pay attention to joint design.


Volvorsport - 16/10/09 at 06:54 PM

www.darrian.co.uk

mosler are made from composite panels bonded together .

using this stuff .

www.teklam.com

my chassis has bonded in kevlar/carbon panels .


StevieB - 16/10/09 at 07:53 PM

But can you bond carbon/kevlar to ali panels to add strength. Not so much strength in the bending of the panel, but a whole ali monocoque laminated to unify the structure properly (if that makes sense?)


Volvorsport - 16/10/09 at 09:25 PM

in my opinion , bonding kevlar/carbon/grp to ali isnt really applicable in DIY situations since the need for temperature control and the surface preparation needed .


time has moved on , there maybe products/applications that would work , but, theres no reason to have carbon/kevlar bonded to ali since the composite is much better at doing it .

if you look at the teklam site , you can indeed get get honeycomb section with ali one side , carbon the other .


andkilde - 16/10/09 at 09:43 PM

You just want t be able to do this:


andkilde - 16/10/09 at 10:46 PM

Have you been following Denis Palatov's progress on the composite chassis for his DP1?

The pages on the composite chassis are in the "past" logs on www.dpcars.net

http://www.dpcars.net/dp1test/dj.htm
http://www.dpcars.net/dp1v8/da.htm
http://www.dpcars.net/dp1v8/db.htm

He's built his second prototype chassis using honeycomb and a thermoset adhesive.

Doesn't look particularly low cost, OK, doesn't look at all low cost but does look exceptionally well thought out.

t


StevieB - 17/10/09 at 03:06 PM

The Teklam products are very similar to those offered by Ayres as used on Trev.

The advantage I see of using the composite panels over ali panels is that once the monocoque is formed, an overall skin of woven roving, carbon or kevlar (budget allowing...) can be added to put some extra strength to the whole thing.

Obviously it will take a serious amount of planning to get it right and I though of starting with a steel speceframe to get the suspension right first, then construct a monocwue to mount everything to.

Of course it's all an exercise in 'because I can/might be able to'....


ceebmoj - 17/10/09 at 04:02 PM

quote:
Originally posted by nasty-bob
Hi

Honeycomb panel was supplied by Technical Resin Bonders and cost about £250 for a 8 x 4 sheet 1/2 inch core sheet.

The carbon/kevlar was from East Coast Fibre Glass Supplies

Cheers

Rob


is who nasty bob used for his compersit shears and it looks like thay have an apropreat product for this ripe of construction


Benonymous - 19/10/09 at 10:35 AM

The Ayres product used in TREV is a honeycomb panel with GRP faces rather than aluminium. They use epoxy resin and a woven mat. A friend who runs a CNC routing business has a few samples.

I think a good approach with this material would be to use it partly as a structural material and partly as a core in your planning.

It would be possible to get a laminate to make a good bond with some surface prep. The choice of reinforcement I would choose would be carbon fibre. Kevlar is an excellent fibre but it is an absolute mongrel to cut and almost impossible to sand. It fuzzes up something chronic and will drive you nuts.

The epoxy GRP honeycomb product is substantially cheaper than the aluminium faced types and I think it would be an excellent choice particularly if you had it cut on a large scale CNC router.

Laser cutting isn't very suitable for aluminium, youre better off cutting it with a rotating tool or a water jet. The GRP stuff could go on a laser but the cost per hour compared to CNC routing is much higher. Any decent kitchen manufacturer would have a CNC router up to the job.

As to design, never feed a load into an unsupported face, only into a supported edge or corner. A good method would probably be to use the HC sheet to make the length of the monocoque and steel to make bulkheads and suspension/engine/steering pickups and structures. The spaceframe components could bolt to lugs or bobbins laminated into the structure.

The biggest downside to this is if you crash it and bust it, you'll have to replace the monocoque part completely.


Volvorsport - 19/10/09 at 05:14 PM

youve just described how the mosler chassis is made .

it uses 1 inch teklam sheet (ally honeycomb with nomex/grp mix) , the inside is laminated with carbon and kevlar , the roof structure is carbon and bonded to the honeycomb chassis .

the rear engine/ suspenion subframe are bolted through bobbins into the chassis as is the front .

the floor section is purposely left long to support the rear subframe as it rests on it , it bolts upwards at the front .

they use large bobbins to bolt everything in .

i cant believe how easy it all looks , a guy with a router and some straight edges could easily make it .


StevieB - 19/10/09 at 06:51 PM

quote:
Originally posted by Volvorsport
i cant believe how easy it all looks , a guy with a router and some straight edges could easily make it .


That has famous last words written all over it!

I agree though - it does look really easy and I guess an easy way to get a template and refine the design is to build a mock up out of ply wood or mdf (I prefer ply as mdf leaves nasty dust). Or maybe makind it using nidaplast depending on the cost of it.

[Edited on 19/10/09 by StevieB]


andkilde - 19/10/09 at 10:02 PM

quote:
Originally posted by StevieB

I guess an easy way to get a template and refine the design is to build a mock up out of ply wood...



Or perhaps even use wood as your "engineered" material. Some fairly interesting sports racers were built with wooden monocoques back in the 60s.

Isn't there a proven wooden seven as well?

Wood is cheap, easy to work with, takes glue well...

t


Benonymous - 20/10/09 at 06:24 AM

Plywood would be the only "engineered" wood product I can think of and compared to using honeycomb as a core, it would be extremely heavy for the thickness. Speaking of which, it's the thickness of the core that increases the strength of the panel.

Frank Costin made a super little car out of plywood, don't remember the model. He was also known to have used plywood for other projects aimed at improving car aerodynamics.

So plywood is definately doable but not as a core so much unless your design aims can be achieved with relatively thin panels.

TREV used 20mm panels that were routed part way through, then bent at the cut, preserving the skin on the outside of the curve/radius. Using simple jigs, the bent panels are held in place while a lamination of reinforcement was laid over the assembly. when it cures, the shape holds nice and simple! You could do a whole centre section of a seven like this.

Properly designed and cut, you might only need two or three pieces to make the tunnel/passenger/sides part of the chassis. It would weigh a heap less than steel tube as well.


StevieB - 20/10/09 at 03:57 PM

A monocoque se7en using this technique should be fairly simple (famous last words...), with my thoughts being to build the cockpit and engine bay as one with subframes front and rear to carry suspension and diff, though I ould extend the floor so that it's fully flat bottomed (and put an F1 style plank on the bottom as a sacrificial layer).

With a bit of planning, you could make it to take boywork from a normal se7en type car (whoever currently supplies locost bodywork...).

I suppose it depends on how much lighter the end result would be compared to a steel spaceframe.

I would only use plywood as a proof of concept and final template, not for the final shssis as it's too heavy.


Theshed - 20/10/09 at 08:24 PM

At least with carbon skins you would not have to have a love affair with rivets.........


A view from above - front pinned on
A view from above - front pinned on


Benonymous - 20/10/09 at 09:56 PM

Very impressive ^^^^^^^^^^^^^^^

What is this a chassis for? The dedication to do all those rivets is admirable. Not forgetting of course, that every hole needs to be deburred prior to seating the rivet. Massive time/effort.


Theshed - 21/10/09 at 10:33 AM

Dedication....there are days when I wished the shed would burn down. Each hole is drill, debur between sheets, drill to size and debur again - countersink for some holes - apply glue - then try and work out why all the holes have moved.....

If I started again I am not sure I would use aluminium. My tub is not that light - circa 100kg. It is probably fairly stiff and will be more so when I add front stays to a roll hoop.

Back to the topic - I think that a carbon honeycombe tub for a 7 style car would be a lot simpler. I think that it has already been done. There is honcombe and honeycombe so beware. It is also used for things like exhibition stands etc and designed to look pretty. Specifying 1/8" cells is a good idea.

It is fairly easy to bond in inserts but even easier to have then bonded in when the panel is made. Yoou would need a jig - a group buy perhaps..

I have not made a great job of load paths at the front of my car and have all sorts of internal reinforcement adding to the weight. The suggestion of tubes from the bulkhead forward would certainly simplify things - that was my plan A and I wish I had stuck with it. The rear ought to be fairly simple though provided that there was a substantial bulkhead behind the seats.

I would make the scuttle structural - a good few layers of CF.


I would make a cardboard model 1/4 - cheeper and more acceptable on the kitchen table. Then a wooden full size model.

I would endorse the suggestion of a cnc router - I did mine by hand...awful!

Given that a measuring error would not be reversable I would gather together all of the bolt on bits before bonding anything.


ceebmoj - 21/10/09 at 12:04 PM

quote:
Originally posted by Benonymous The choice of reinforcement I would choose would be carbon fibre. Kevlar is an excellent fibre but it is an absolute mongrel to cut and almost impossible to sand. It fuzzes up something chronic and will drive you nuts.


As i under stood it in treve the Kevlar is used to make the floor and interior puncture resistant. Will CF offer the same benefits?

also does the GTM use CF / kevlar in its tub? Looking at a number of resources on line it looks like it would be more efisent to use a mix of fabrics.


Benonymous - 22/10/09 at 02:37 AM

Kevlar would definitely be the best choice for puncture resistance and if you just cut a section that had no need to be trimmed and put it in the layup, it wouldn't provide you with too much grief. CF would not give you the same properties as Kevlar but I don't see it as being a major problem if you had a 20mm panel and CF reinforcement it would be more puncture resistant than say an aluminium floor.


ceebmoj - 22/10/09 at 12:16 PM

hi,

this is my attempt at some cad of a tub all 20mm core to be covered in resen and fabric what do people think?

Its a 60s inspired single seater. the curved side panels will be lade up seporatly then bonded on to the tub

compersit tub
compersit tub


[Edited on 22/10/09 by ceebmoj]


Theshed - 22/10/09 at 01:13 PM

Is that rear engine - feet through the window of the middle bulkhead?

I would try to design out the small fillet on the inside bottom.

Where are your suspension loads going?

One thing I think I got right is a fully stressed seat base - a fully stressed back would help.

Are you aiming for a semi stressed engine but a fully stressed gearbox? If so you need to think how you will attach your engine plate/adaptor. You can have inserts bonded in when the panel is made.

A 60's style car gives you a nice choice of transaxles if you go for engine in the back - Hewland Mk 9 boxes are very reasonable.


ceebmoj - 22/10/09 at 03:04 PM

quote:
Originally posted by TheshedIs that rear engine - feet through the window of the middle bulkhead?


Yes that the idea.

quote:
Originally posted by Theshed
I would try to design out the small fillet on the inside bottom. 


I put the filet in there so that there would be plenty of contact patch to glue the triangle to the floor with like the trev project. I also left the end cap of to show the detail

quote:
Originally posted by Theshed
Where are your suspension loads going? 



I have marked the rear suspension points on the new picture

compersit tub suspention point
compersit tub suspention point


quote:
Originally posted by Theshed
One thing I think I got right is a fully stressed seat base - a fully stressed back would help. 


I have been trying to incorporate a stressed back but want to put my fuel tank between the seat back and the rear bulkhead so was going to have another bulkhead with a window in it so that I can still get access to the fuel tank by removing an alloy panel.

quote:
Originally posted by Theshed
Are you aiming for a semi stressed engine but a fully stressed gearbox? If so you need to think how you will attach your engine plate/adaptor. You can have inserts bonded in when the panel is made. 

A 60's style car gives you a nice choice of transaxles if you go for engine in the back - Hewland Mk 9 boxes are very reasonable.


Not sure if I will go stressed box i'm considering using a bike engine and just hanging a car diff out the back having a plate the diff mounts to and the rear suspension pick up points are on that its then bolted on to the Tub

[Edited on 22/10/09 by ceebmoj]


StevieB - 23/10/09 at 09:20 PM

That looks good to me - would you do the outer skin with honeycomb as well?

I need to get a copy of CAD in order to plan out my own design! Looks much neater than my hand sketches


ceebmoj - 24/10/09 at 04:52 PM

I was thinking of just laying up the side panels out of two layers of cloth and resen with a gell cote and then bonding them on to the tub.

my uprights are out being machined at the moment so after that I will make up a MDF tub and do a build up with that


Ferg - 24/10/09 at 06:45 PM

quote:
Originally posted by ceebmoj
also does the GTM use CF / kevlar in its tub?

No.
The Libra and Spyder tubs are purely GRP with the exception of two 2" steel plates which are the upper seat-belt mounts. Otherwise......nothing but GRP and the Gelcoat.


Uryen - 25/10/09 at 08:43 AM

quote:
Originally posted by Theshed
Are you aiming for a semi stressed engine but a fully stressed gearbox?

Could you explain what this means?
Does it mean the engine and box are part of the frame structure like in some extruded aluminum bike frames?


StevieB - 25/10/09 at 09:29 AM

quote:
Originally posted by Uryen
quote:
Originally posted by Theshed
Are you aiming for a semi stressed engine but a fully stressed gearbox?

Could you explain what this means?
Does it mean the engine and box are part of the frame structure like in some extruded aluminum bike frames?


Pretty much the same principle, yes.

If you have a look at the beack end of a formula fors, the gearbox carries the rear suspension and the whole rear end is supported by the engine. Rescued attachment ff engine.jpg
Rescued attachment ff engine.jpg


Theshed - 1/11/09 at 10:25 AM

Yup that's what I meant ^^ What I meant by "semi stressed" was that the engine would be solid mounted but braced by some part of your composite structure.

I would think about a nice big plate across the back of the structure to tie everything together.

Excellent plan to start with the uprights and work inwards.

If you look on GT40s.com in the build logs you will see somebody building a carbon tub using mdf to make moulds etc

John


aitch - 2/12/09 at 07:53 PM

composites or 2 layers of structural material (aluminium, fibreglass carbon fibre whatever) sandwiching a spacing material are far stronger and lighter than the moulded fibreglass type construction.

using woven glass and epoxy resin sandwiching foam will be far cheaper than the honeycombe materials although forethought and planning is paramount as reinforcing needs to be built in rather than just bobbins as the foam material will crush more readily.
Its also easy to produce curved components.
The processes are very well documented in the kit aircraft market, i believe the techniques originally pioneerd by Burt Rutan and used in many of the worlds most popular and numerous homebuilt aircraft

aitch

p.s. new here and thinking aloud please be gentle with me

[Edited on 2/12/09 by aitch]

[Edited on 2/12/09 by aitch]


aitch - 3/12/09 at 12:03 PM

Ive re-read this whole thread and the links

the Nidaplast cores are very interesting, i would be more concerned about heat resistance than strength, however these cores would make possible a locost composite chassis that is TRULEY LOW COST more expensive aluminium honeycombe could be used in the areas that need the heat resistance, id opt for woven glass and epoxy over most of the structure to keep costs reasonable, although carbon and or kevlar could again be used in some areas either for strength, puncture resistance or asthetics
http://www.cfsnet.co.uk/acatalog/Nidaplas_Polypropelene_Core.html they also provide links to use and application of the core material which would be invaluble in planning the build

The Palatov Chassis uses composites to create a series of box sections, a technique which followed will create incredibly strong and rigid structures

aitch
im going to use some of this time to come up with a concept

[Edited on 3/12/09 by aitch]


aitch - 3/12/09 at 12:08 PM

as it says good pic of the palatov chassis showing the box sctions created




aitch Rescued attachment dp1017.jpg
Rescued attachment dp1017.jpg


aitch - 3/12/09 at 02:36 PM



also been talking with the technical Guy at CFS (Alex) and on the choice of resins and cloth his opinion is that polyester resins and chopped matt would be more than adequate, if so the costs would fall drmatically compared to woven cloth and epoxy resins.. defonately some structural tests needed..
Im only at the thinking about it stage although anyone ready to get started i am sure Alex at CFS will be more than helpfull
aitch

[Edited on 3/12/09 by aitch] Rescued attachment Kit Car.JPG
Rescued attachment Kit Car.JPG


kennyrayandersen - 3/12/09 at 03:27 PM

I’m a stress analyst working in the aerospace field with nearly 25 years of experience – maybe half of that (or better) is working with composites; so, when I see a topic like this I feel a bit compelled to chime in. You can believe what I will tell you or not – I’m not trying to sound arrogant, and there are people who know more about composites than me, but as a percentage of population, not so many.

quote:
Originally posted by smart51
I love it when people step into this bear trap. It's even better than CEC vs BEC.

Basically, Ali is too brittle for the vibrations seen by a space frame chassis and will snap before too long. It may be all right for a car you're going to use once (say you're entering the lemans 24hr) but not for regular use.


This of course is utter nonsense since composites are used for all kinds of airplanes, sporting equipment, and military applications where there is a lot of abuse and vibration. In fact, one of the primary incentives for using composites is its excellent fatigue properties. It’s a matter of design and keeping the strains below the level that any delaminations due to assorted impacts whether they be VID (visible impact damage) or BVID (barely visible impact damage) will propagate when loaded ‘normally’. Brittleness has nothing to do with anything unless you are talking about a crash where the material is expected to fail catastrophically, and where graphite will shatter quite spectacularly.

quote:
Originally posted by Benonymous

I think a good approach with this material would be to use it partly as a structural material and partly as a core in your planning.


Where to use what requires a firm understanding of the various failure modes of the composites. The solid, as well as the sandwich material, can’t be thrown around willy-nilly and expect it to come out OK in the end. Its really too complicated to go into in-depth in a paragraph or 2 but suffice it to say that an engineer, or long-time composite specialist should be involved in anything more than a scooter.

quote:
Originally posted by Benonymous

It would be possible to get a laminate to make a good bond with some surface prep. The choice of reinforcement I would choose would be carbon fibre. Kevlar is an excellent fibre but it is an absolute mongrel to cut and almost impossible to sand. It fuzzes up something chronic and will drive you nuts.


Kevlar is horribly weak in compression and so not used for any strength-type application but to line the inside of the cabin to protect the occupant from splintering graphite (and it will splinter when crashed). It is great in tension and will strain like crazy, so it does have its place.

quote:
Originally posted by Benonymous

The epoxy GRP honeycomb product is substantially cheaper than the aluminum faced types and I think it would be an excellent choice particularly if you had it cut on a large scale CNC router.


Either is quite expensive, and I would be stunned if an aluminum facesheet/aluminum core were cheaper than graphite facesheets with a NOMEX honeycomb core. Seriously, graphite is really expensive and even fiberglass facesheets (which I think you are referring to) on whether a foam or honeycomb core is still quite expensive.

quote:
Originally posted by Benonymous
Laser cutting isn't very suitable for aluminum; you’re better off cutting it with a rotating tool or a water jet. The GRP stuff could go on a laser but the cost per hour compared to CNC routing is much higher. Any decent kitchen manufacturer would have a CNC router up to the job.

As to design, never feed a load into an unsupported face, only into a supported edge or corner. A good method would probably be to use the HC sheet to make the length of the monocoque and steel to make bulkheads and suspension/engine/steering pickups and structures. The spaceframe components could bolt to lugs or bobbins laminated into the structure.

The biggest downside to this is if you crash it and bust it, you'll have to replace the monocoque part completely.


The whole manufacturing process is complicated and expensive. And as they say, the devil is in the details. The hardest part is joining it all together which would include specialized inserts, brackets and fittings that were specially designed for the application. Machining honeycomb is a beotch, to say the least, but it is an essential part of the process. If can be ordered from Hexcel expanded and machined, but the cost is simply staggering.

quote:
Originally posted by aitch
as it says good pic of the palatov chassis showing the box sctions created



aitch


That chassis is simply nonsense and I wouldn’t get into it to go around the block. It shows a complete lack of understanding of load paths and proper application of composites in general. This is a poster child for why you shouldn’t do this at home! I know this is going to come off as overly harsh, but I believe the intent for most folks is to drive these on the road, no? The last thing you want is to be a danger to yourself and those around you. I was simply gob smacked when I took a look at that.

NOW, let’s ask a simpler question… Why? Generally the only reason that we use composites is that we REALLY want to save some weight. We are willing to pay huge costs and manufacturing penalties to do so; however, it should be noted that these applications are for very high-tech nearly money-is-no-object projects that have very, very deep pockets. Think formula 1 or space shuttle.

My Eurocopter buddy was telling me that though they have the capability and ability to make a nearly all-composite helicopter (including RTM), they are looking to use primarily aluminum on their next generation product because the cost, effort and trouble of composites are just not worth it (this from people who really know how to do it from beginning to the end). If you are going to build an F-22 there is no way to avoid using composites. If you are going to build a Locost, or any other home-built vehicle, IM not-so-humble O you are barking [mad] up the wrong tree!

(takes a deep breath)

Its simply silly.

[Edited on 3/12/09 by kennyrayandersen]


aitch - 3/12/09 at 03:47 PM

kennyrayandersen
obviously WAY more experienced and knowledgeable than i, and i always appreciate imput especially when its negative as the posibble outcomes are horrific..

ive no idea what kind of stress analysis the petrov people have done, somehow the box sections in the chassis tub just "LOOK" right when compared to wing spars and cockpit tubs of the homebuilt aircraft, however possibly light on structure and load paths around the suspension mounting areas ????

aitch


twybrow - 3/12/09 at 03:57 PM

Some great points there Kenny - hopefully I am one of the few who does know enough to add to this - you have raised some very valid arguments. I typed this and walked away, and you had covered most of the points I was going to touch on, so I won't repeat too much....

aitch - If you are being advised to use CSM and polyester, then the design of your chassis must be pretty pants! What is the point of going for a composite chassis, if you don't fully utilise the benefits of composites. If you use CSM, you will add a significant chunk of weight and reduce the mechanical performance considerably (CSM is only 1/3 the stiffness of a unidirectional material).

you seem to have designed/built that as if you were building out of an isotropic material (which CSM effectively is in X and Y directions), but composites come into there own when you consider them as the anisotropic materials they are. Consider you load paths, and consider youravaiobale materials, mabnufacturing techniques, inspection methods etc.

Are you doing this as a learning exercise, or to get the lightest, stiffest chassis you can? That will make a big difference to how you can progress from here....


aitch - 3/12/09 at 04:19 PM

quote:
Originally posted by twybrow
Some great points there Kenny - hopefully I am one of the few who does know enough to add to this - you have raised some very valid arguments. I typed this and walked away, and you had covered most of the points I was going to touch on, so I won't repeat too much....

aitch - If you are being advised to use CSM and polyester, then the design of your chassis must be pretty pants! What is the point of going for a composite chassis, if you don't fully utilise the benefits of composites. If you use CSM, you will add a significant chunk of weight and reduce the mechanical performance considerably (CSM is only 1/3 the stiffness of a unidirectional material).

you seem to have designed/built that as if you were building out of an isotropic material (which CSM effectively is in X and Y directions), but composites come into there own when you consider them as the anisotropic materials they are. Consider you load paths, and consider youravaiobale materials, mabnufacturing techniques, inspection methods etc.

Are you doing this as a learning exercise, or to get the lightest, stiffest chassis you can? That will make a big difference to how you can progress from here....


just to clarify, the black tub in photo's is the palatov chassis as per link's on page one of this thread, the sketch is just a very quick ilaberation on using a similar meathod for a mid engined BEC..

my initial thoughts were to use uni and bi cloth with epoxy resins, the supplier of the honeycombe material (who's experience is with marine applications where weight is not so much of an issue) suggested CSM... i entirely see your point about the additional weight required for strength in this application. particularly givven the high ratio of resin when using CSM

so undoubtedly woven cloths and epoxy would be the way to go "if at all" i am in the very early stages and have to say the idea of composites apeals to me,,,,, so does getting it right and being safe..

It could well be that i will have no option but to take a trqaditional steel frame route as i have no access to the technical expertise needed to realise a composite chassis nor the wealth for many prototypes

aitch


twybrow - 3/12/09 at 04:34 PM

Thanks for the clarification.

Don't be too hasty to give up, but just have a think as to what you are aiming to achieve. You could gain plenty of knowledge by going to your local library, and start reading! There are sooo many books out there on composite design.

Can you use a CAD package? FEA? Have you ever designed/manufactured anything in composite before?


aitch - 3/12/09 at 04:39 PM

quote:
Originally posted by twybrow
Thanks for the clarification.

Don't be too hasty to give up, but just have a think as to what you are aiming to achieve. You could gain plenty of knowledge by going to your local library, and start reading! There are sooo many books out there on composite design.

Can you use a CAD package? FEA? Have you ever designed/manufactured anything in composite before?


Seriously now dont laugh
model aircraft, and resesarch into homebuilt composite aircraft untill my flight training stopped 7 years ago

im not givving up on the idea and have loads of time while saving for the build

ive never used a cad package but have experience in various none 3d design packages, if i could pick up a cad package that was good enough and a reasonable price then in time i would be able to use it

aitch


aitch - 3/12/09 at 05:37 PM

purely a proof of concept

http://www.tech.plym.ac.uk/sme/acmc/download/chass.pdf

aitch

p.s. looking forward to months of reading and learning


twybrow - 3/12/09 at 06:44 PM

^ That was my dissertation supervisor back in the day! He knows his onions!

Some more reading for you:

Here

Here

You'll like this one

Here

[Edited on 3/12/09 by twybrow]


aitch - 3/12/09 at 10:02 PM

quote:
Originally posted by twybrow
^ That was my dissertation supervisor back in the day! He knows his onions!

Some more reading for you:

Here

Here

You'll like this one

Here

[Edited on 3/12/09 by twybrow]


yes some very interesting reading and i do like that one !!!!!!

the space frame is interesting although i will have to read more to understand the mulding process, and i will be reading more

aitch


kennyrayandersen - 4/12/09 at 12:02 AM

I wrote that quite late at night, so maybe I was a bit grumpy; however, the points were pretty much valid. The problem with the Palatov chassis is primarily that the vertical panels he has in the back don’t line up with anything else; so, when the load gets to the bulkhead it’s reacted as bending, which is the least efficient possible way to do it. The reason one adds honeycomb core to two thin facesheets is to make the panel more stable [it’s still got to carry the load]. The catch there is that when the load gets to the edges of the panel, you have only these 2 thin facesheets to react the load. In aluminum, the bearing allowable is around 120-130 KSI ultimate, so you can make the core tapered at the edges and form the aluminum facesheets and get away with having thin edge members.
Conversely, the composites are much worse in bearing. Without a lot of sophisticated software, you could use a value of 60 ksi (for graphite epoxy which has been cured in an autoclave) which is around half of aluminum. Therefore, with the composite facesheets you need to [typically] add extra plies around the boarders of the panel for bearing strength – the core still typically has to be machined, or purchased as a pre-machined detail to be made into a composite assembly. I took a quick look through some of the links that twybrow posted and you’ll not that some of those techniques will only work with aluminum core and facesheets, since they require the ability to crush the core or yield the facesheets in bending – nether of these are permissible with graphite or fiberglass as it will catastrophically fail the panel. And actually it’s not such a great idea in aluminum either as the crushed core just isn’t the same (pre-failed?).

Generally, you note that most composite race cars and such use the material to make a monocoque-type tub of sorts where the panels can be loaded primarily in shear. The trick, then, is to get the loads which are introduced at singular points in to broader shear loads in the composite panels. This is done with the fittings and such that I mentioned in my earlier post. It’s not that it can’t be done of course, but you will likely need your own machine shop or be prepared to ante up to have the work done ‘out of house’.
Let’s say a decently designed and fabricated steel locost chassis weighs what? 85-100 Lb? Let’s say for the sake of argument that you’re a über-genius such as myself (tongue-in-cheek). Maybe I could get the weight down to 65-75 Lb using finite element analysis and what I know about composites (which as any actual genius knows – no one can know that much about anything).

What did I actually ‘pay’ to make this happen.
1. Time, initial -- extra design and analysis time
2. Money -- from something nearly any bloke with a paper route could buy to something that has to be funded by NASA – this is just with regard to materials
3. Time , manufacturing – briefly, labor INTENSIVE. What was a couple of weekends has now gone on to a couple of months… or years
4. Facilities – where you going to get that autoclave?, or will you ‘settle’ for the lower allowables from a room temperature system? McLaren bakes, enough said.
In summary, to save 25 pounds, you sold your soul and you left testicle to the devil (who was in the details). A body is a different story altogether, but we are talking chassis here. If a steel chassis is good enough for the £39,275 pound Caterham Superlight 500 (and might I add super-fast), it’s good enough for yours!

[Edited on 4/12/09 by kennyrayandersen]


Benonymous - 4/12/09 at 06:00 AM

Thanks for your input Kenny. There's no doubt that the good old steel space frame is pretty hard to beat. The weight comparisons you made are really quite startling if they are accurate. So its a few hundred quids worth of steel up against thousands of quids worth of composites.

Still the garage culture will carry on and the amateurs will eventually figure out what value they can get from these sorts of materials. The main thing that's lacking at the present time is readily available information on using composites.

By contrast, there's no end of info about steel and welding. We will have to wait until carbon fibre is used to make commonly used items and is no longer a black art.

Personally, I'd never consider making a chassis out of composite unless I'd prototyped it in some way and possibly done some FEA in SolidWorks or something similar.

Without people experimenting with the stuff, the average bloke has no hope of using carbon. I'd happily build something out of it if I thought I could make it safe. I already have a vacuum bagging rig and soon I hope to have a medium sized CNC router to make moulds and plugs with.


kennyrayandersen - 4/12/09 at 07:14 AM

quote:
Originally posted by Benonymous
Thanks for your input Kenny. There's no doubt that the good old steel space frame is pretty hard to beat. The weight comparisons you made are really quite startling if they are accurate. So its a few hundred quids worth of steel up against thousands of quids worth of composites.

Still the garage culture will carry on and the amateurs will eventually figure out what value they can get from these sorts of materials. The main thing that's lacking at the present time is readily available information on using composites.

By contrast, there's no end of info about steel and welding. We will have to wait until carbon fibre is used to make commonly used items and is no longer a black art.

Personally, I'd never consider making a chassis out of composite unless I'd prototyped it in some way and possibly done some FEA in SolidWorks or something similar.

Without people experimenting with the stuff, the average bloke has no hope of using carbon. I'd happily build something out of it if I thought I could make it safe. I already have a vacuum bagging rig and soon I hope to have a medium sized CNC router to make moulds and plugs with.


To be sure the general knowledge about composites is much less advanced, but the conundrum is really quite a bit more complex than that. It’s not just knowledge that is holding people back from using composites – there are a lot of factors. Just because you can do a thing, doesn’t mean you should do a thing. In any given application there is always trade-offs and compromises. Really, this is the art of engineering. On any given airplane, for instance, there can be found aluminum, steel, graphite, fiberglass, various adhesives, titanium, and even Velcro! The right material is the right material regardless of whether it seems all cool and high-tech.

1) I’m fully capable of designing and even making composite parts; however, I’m not even thinking about it. Why? The biggest hurdle is cost. I’m mean as a cost-no-object exercise in I-told-you-so, it makes a great demonstration, but I have better things to do with my time and money (as you get older you think about things a little differently (you realize your time isn’t infinite and you’re going to need a fair pot-o-coin to get you through the old years)). If you are sending something into space it might be worth a few thousand quid to save a kilogram of weight; if you are building a ship – not so much.

2) The second hurdle is manufacturing. Even the above average bloke isn’t going to have, or even have access to, an autoclave. Buy the time you take a knockdown for the room temperature cure resin system and the fact that without it you will either gain weight due to excess resin (since you will have to use a wet layup as opposed to a prepreg (forget RTM since you’d have to make all the tools before you could make the parts and the upfront costs are to be considered not insignificant (I actually have a friend that made a spar in his garage this way, so yes, it can be done but…)), or have an increase in porosity, it’s even a bit less attractive. Then there is the labor. Since you don’t have a multi-million dollar tape layup machine like Boeing, you are going to be regulated to hand cutting and laying up the composite.

If you are really dead set on having some ‘fun’ with composites, I suggest you start messing around with some fiberglass/epoxy and some rigid foam (the kind they use in house insulation – pink or blue EPS -- Expanded Polystyrene). Start by making a hood, or maybe a se7en side door, or a se7en boot cover, or maybe a dash (this would be great made out of glass or graphite facesheets with a foam core). As you’ve noted and as I’ve pointed out a chassis is reserved for some highly trained and deep-pocketed ‘teams’ of people to do. A body, body panels, floors, or panels etc. are more realistic applications that a DIY guy could expect to do and get good results. I’m not saying don’t do anything out of composites, but whatever you do try and be sensible and realistic about it.


aitch - 4/12/09 at 09:40 AM

Ok what sort of cost am i looking at to get a CAD package with FEA and suggestions or recomendations as to which one i should go for

aitch


aitch - 4/12/09 at 10:35 AM

ok had a look around at some cad packages for medelling and FEA tools to go with them,,

unfortunately way out of my price range and for a one off prototypes of chassis parts and testing to destruction would be more cost effective (along with advise from someone in the knw to get basic design concept)


twybrow - 4/12/09 at 01:16 PM

You can pick up dodgy copies of most on the internet.

I used ProEngineer/ProMechanica, but it isn't the best package out there in terms of interfacing witother packages. CATIA seems to be the most widely used CAD package, and NASTRAN seems to be used extensively for FEA in the composites world.


Alan B - 4/12/09 at 03:47 PM

FEA......another hot potato

Without a good deal of conventional experience (read years of calculators and number crunching) FEA is at best just a rough guide. If, of course, you can even get to work in the first place...most are very un-user friendly.

It's a bit like saying being able to use MS word makes you an author......just because you can churn out some nice coloured plots doesn't make you a stress analyst.

Just a little rant in general, certainly not having a go at anyone in particular.


Volvorsport - 4/12/09 at 04:30 PM

www.darrian.co.uk

what do you make of these then kenny ?


kennyrayandersen - 4/12/09 at 11:43 PM

FEA packages

First, not many packages that are good at designing and solid modeling are good at FE modeling and vice versa. All of the people around me that design use CATIA V5, which will cost you both your stones and one kidney. The package that I use for FE work is NASTRAN, which is obscenely expensive and which also requires a preprocessor – I use PATRAN (both are MSC products) as the preprocessor (you can also use FEMAP, which is a ‘bit’ cheaper, that some guys seem to like (my German colleague seore by it)). Each of these tools are extra ordinarily powerful, but much like a gun in the hands of a child can do a lot of damage without proper experience and training. THere are other good packages like COSMOS etc. but the good ones, have good support and are feature laden, so they are quite expensive, especially by DIY at home standards.

There are cheap packages, which are mostly crap, and as noted none of them are particularly easy to use. I’ve been using NASTRAN output for going on 25 years in one form or another and have been using PATRAN for maybe 15 years. The cheapest way to get to use and learn one of these packages is by becoming a student (as stated earlier).

I wish I could tell you there is a magic button to push, or that there was some shortcut to knowledge, wealth and the attraction of stunning hotties, but that would make me and instant liar. Even if I had a fiver for every time I saw a dysfunctional, poorly constrained and loaded and utterly worthless [and therefore dangerous] FE model, you could just call me the Count of Monty Christo! Keep in mind these worthless models are from people with 4-year (or more) engineering degrees and some experience – it really is that hard to do well (regardless of the advert on the FE package website!). I really don’t think you are 1) listening, and 2) being realistic about what you are going to be able to do. If you could only channel that determination into something that will further your understanding of the underlying principles, you’d be virtually unstoppable. I’m not trying to dis you, make you mad or discourage you in any way. I am, given my background, trying to save you some scratch and give you some valuable [and free] advice – this is from someone that, though not knowing much, knows a great deal more about this than most.

If you really can’t be dissuaded from diving into the pool head first, then you should consider finding a proven design and then copying it. This might require a bit of destructive testing, but even that would be cheap compared to the path you are contemplating.

Does anyone think I’m being overly harsh here?


Darrian
It looks fine from the outside (though maybe a bit dated?). I was hoping to take a look at the chassis, but couldn’t find any pictures. In the description, it said it used a monocoque tub, which I said in one of the earlier posts was generally the only way to go. Also, I read that March was involved, and they have made a ton of race cars, so it probably fine structurally. Note though that my other points are all contained in this car. It’s mostly used for racing (as opposed to the road), it’s expensive – well sorted ones were being offered at 30-40K£ which is quite apart from the lo-cost concept. Even a roller was 20K£! This just backs up what I was saying:
1) Done by people who know what they are doing
2) Expensive
3) Difficult to produce
4) Used mostly for racing where weight-savings are critical and worth the additional cost (the small difference in weight doesn’t mean so much for a standard road car. Additionally, the robustness (due to increased weight) is actually a desirable thing for a road-going vehicle (early Lots 7’s for instance were very light-weight but broke easily and wore out quickly – neither of which would be a plus on a street car).

[Edited on 4/12/09 by kennyrayandersen]

[Edited on 4/12/09 by kennyrayandersen]


richmars - 5/12/09 at 11:01 AM

kennyrayandersen, too harsh? no, just realistic. Not that I know 1% of what you do re designing a composite chassis, but I have worked with FEA and CFD packages so I know just how easy it is to get very pretty pictures which show what you want to see, instead of what's going to happen!
I've also made a few carbon bike frames just for fun and to learn about the process, but they're all sitting in the loft because I wouldn't risk using them too much for the very reasons you've mentioned. If I had an autocleave, knowledge, time to make 10's of prototypes, a test lab etc I may be able to make something I would trust but not something I did in the garage.
Carbon is a great material, and I love making things in it, but I limit it to non critical parts, like dash boards (did that about 8 years ago) and mountain bike mud guards. On a kit car there are loads of parts that you can do in carbon which won't kill you (or someone else) if they fail.


scootz - 5/12/09 at 11:21 PM

quote:
Originally posted by kennyrayandersen

quote:
Originally posted by smart51
Basically, Ali is too brittle for the vibrations seen by a space frame chassis and will snap before too long. It may be all right for a car you're going to use once (say you're entering the lemans 24hr) but not for regular use.


This of course is utter nonsense since composites are used for all kinds of airplanes, sporting equipment, and military applications where there is a lot of abuse and vibration. In fact, one of the primary incentives for using composites is its excellent fatigue properties. It’s a matter of design and keeping the strains below the level that any delaminations due to assorted impacts whether they be VID (visible impact damage) or BVID (barely visible impact damage) will propagate when loaded ‘normally’. Brittleness has nothing to do with anything unless you are talking about a crash where the material is expected to fail catastrophically, and where graphite will shatter quite spectacularly.


Am I missing something? Smart51 is talking about aluminium (ali) being too brittle for a spaceframe chassis and your reply stating it's 'nonsense' relates solely to composites...


kennyrayandersen - 6/12/09 at 05:24 AM

quote:
Originally posted by scootz
quote:
Originally posted by kennyrayandersen

quote:
Originally posted by smart51
Basically, Ali is too brittle for the vibrations seen by a space frame chassis and will snap before too long. It may be all right for a car you're going to use once (say you're entering the lemans 24hr) but not for regular use.


This of course is utter nonsense since composites are used for all kinds of airplanes, sporting equipment, and military applications where there is a lot of abuse and vibration. In fact, one of the primary incentives for using composites is its excellent fatigue properties. It’s a matter of design and keeping the strains below the level that any delaminations due to assorted impacts whether they be VID (visible impact damage) or BVID (barely visible impact damage) will propagate when loaded ‘normally’. Brittleness has nothing to do with anything unless you are talking about a crash where the material is expected to fail catastrophically, and where graphite will shatter quite spectacularly.


Am I missing something? Smart51 is talking about aluminium (ali) being too brittle for a spaceframe chassis and your reply stating it's 'nonsense' relates solely to composites...

Mea Culpa – I did think he was talking about composites; I got confused because this thread was about a composite chassis and there was another thread about aluminum being used, which I almost responded to, but didn’t get around to. Maybe there should have been one specifically dealing with material selection, but regardless, my response is valid for both cases.

It’s not vibration per say that is why using aluminum MAY not be a good idea for a street frame, but rather the repeated loading at levels that will cause fatigue cracking. Vibrations, and especially ones at a higher frequency [like that of the engine] than the natural frequency of the chassis, tend to be at a lower level than the fatigue limit and therefore don’t do the damage. What more likely fatigues a frame are the repeated input of high levels of applied torque from the engine as well as the repeated g-loading of just driving it down the road (those loads due to the weight of the occupants and the engine/transmission etc.) and how those loads are reacted buy the chassis (as points near the attachment of the A-arms and the A-arms themselves). If the Aluminum chassis is designed to keep these stresses below the fatigue limit (basically designing for infinite live) then there is absolutely no reason you couldn’t do the chassis out of aluminum. It may not be much lighter in the end, and it will certainly be more costly, but those are entirely different, though not irrelevant, issues.

Perhaps Smart51 used the term vibration loosely, not realizing that is actually has a more specific meaning and thus a bit different implied meaning that what he was thinking. When we talk about vibration, we generally are referring to smaller oscillatory loads that are superimposed onto some type of general load cycle. For instance, a car can experience up to 3g down load (due to say a dip in the road – 3g level loading is a standard on-road load factor used in sizing over-the-road equipment)). There may be some vibration or whatnot that might be superimposed on top of that (but not usually). A bracket might be sized for vibrations (like on an alternator, for instance), but I’m thinking for a chassis it is load cycles and not vibrations. This 3g loading would be some type of ultimate load case – this just doesn’t happen that often (I may have experienced this only once as a 15 year old driver while trying traverse a ditch straight on… ouch!). Fatigue loading would be maybe 1.25g or something – I’m not a car guy, I work on airplanes and helicopters, so I’d have to do a little research to know exactly what the industry is using. Anyway, the fact that almost all of the worlds planes and helicopters are made from aluminum suggests that vibrations are not the problem, but rather the load level that is repeated that is at a level that does fatigue damage that is the issue.

The one thing we do know without having to get all scientific is that steel is much more fatigue tolerant that aluminum. Steel still has fatigue issues (early 7s were plagued by cracking chassis, but to be fair they were race cars!), but it is more tolerant. Therefore, I guess the bottom line is that it’s easier to make the frame out of steel, and more especially it’s easier for the home fabricator to get from here to there using steel for the chassis.


Benonymous - 6/12/09 at 09:22 AM

One of the links in a previous post went to a .pdf file with among other things in it the chassis of the Lotus Elite. Now while I'm not going to pretend that it was built in a felt-roofed shed in the back yard of a semi in Finchley I'd venture to suggest that it was designed with a fairly basic set of tools where load and stress was concerned. One stand out characteristic was that the suspension mountings used to fracture, not good... However let's say for the sake of example that I decided to build a Lotus Elise chassis from Carbon fibre and epoxy resin, using the appropriate layup thicknesses etc. Surely this would produce a superior wet layup than glass and polyester? If this is so, then I would have a stiffer, stronger chassis than an Elite? So maybe I wouldn't have to limit my carbon fabrications to cosmetic parts?
I respect your knowledge and understanding kenny but ultimately, experimenters will do things that engineers would never dream of and maybe even put themselves at risk in the process. I feel that this is the way of advancement. Without these risk takers there would be no kit cars or kit planes, ultralights or even the superb Britten V1000, made by a complete amateur in the field of engine and motorcycle chassis design. Sure, John Britten could have crashed and died on the first corner but thanks to a bit of judicious over engineering, the concept survived and went on to be one of the most heroic efforts in motorcycle racing of all time. Any engineer would have told him he was loopy and to give it all up and grow Petunias!




What a dull world it would be if he had eh?


Theshed - 6/12/09 at 11:06 AM

I love this thread. It is fantastic when the internet brings together really useful knowledge and debate and avoids unnecessary unpleasantness just because the contribtutors disagree.

Please keep posting Kenny experience like yours is priceless. I doubt that you will put off us loonies that depart from the steel spaceframe (that is clearly the obvious amateur choice) so any tips you can give to save us from our own insane choices.

I have made all of my own inserts for my (aluminium) honeycomb chassis. All pure guesswork. I would love to see how it is done in the aircraft industry. Any photos anyone?


kennyrayandersen - 6/12/09 at 01:41 PM

quote:
Originally posted by Benonymous
One of the links in a previous post went to a .pdf file with among other things in it the chassis of the Lotus Elite. Now while I'm not going to pretend that it was built in a felt-roofed shed in the back yard of a semi in Finchley I'd venture to suggest that it was designed with a fairly basic set of tools where load and stress was concerned.

One stand out characteristic was that the suspension mountings used to fracture, not good... However let's say for the sake of example that I decided to build a Lotus Elise chassis from Carbon fibre and epoxy resin, using the appropriate layup thicknesses etc. Surely this would produce a superior wet layup than glass and polyester? If this is so, then I would have a stiffer, stronger chassis than an Elite? So maybe I wouldn't have to limit my carbon fabrications to cosmetic parts?


The point being that those basic tools include basic engineering skills; which even when present is no guarantee there won’t be trouble – this is why even the largest, most sophisticated technology companies in the world have to test after the initial engineering work is done. Sure, graphite will give both a stronger and lighter structure than fiberglass, since it has a higher modulus, but basically any material short of peanut butter can be made to work – some materials are better than others. I saw a 100 year old race car powered by an enormous antique airplane engine. It is still being raced and the frame is made from Ash (the original frame!)!

quote:
Originally posted by Benonymous I respect your knowledge and understanding kenny but ultimately, experimenters will do things that engineers would never dream of and maybe even put themselves at risk in the process. I feel that this is the way of advancement. Without these risk takers there would be no kit cars or kit planes, ultralights or even the superb Britten V1000, made by a complete amateur in the field of engine and motorcycle chassis design. Sure, John Britten could have crashed and died on the first corner but thanks to a bit of judicious over engineering, the concept survived and went on to be one of the most heroic efforts in motorcycle racing of all time. Any engineer would have told him he was loopy and to give it all up and grow Petunias! What a dull world it would be if he had eh?


You make is sound like engineers are all a bunch of chicken $hit$, which may well be true! but I believe it to be only partially true at best. Lilienthal, was a scientist and also a great glider experimenter in the early [pre Right brother] flight days. He, of, course, augured in and was no more. Mark Donahue also comes to mind; he was an engineer and very successful race car driver who unfortunately met an untimely demise racing in his own creation as well. So even IF the knowledge is there, giant screw-ups can still cost you big time. I’ve got not a problem in the world with people trying any random, dangerous and even crazy concept; the only slight hiccup in the natural selection process however is that the ‘weak’ some times take out the strong in the act of removing themselves from the gene pool. Usually this is preceded by the expression “Hey y’all, watch this”.

The post of the motorcycle contains just a bit of irony as the riders of such are frequently referred to as organ donors. Hmmm…

Don’t misunderstand what I’m trying to say here. I’m not saying don’t do it, I’m saying educate yourself before you do (not you in particular, but anybody wanting to go off and be extra ‘creative’); do that the best you can. Also, understand that you might be doing just to say you did i.e. it not the best way to do it even though it CAN be done that way. Last time I check the UK was still free enough that you can spend you hard earned cash on whatever your little heart desires; people have gotten less for more. Knowing what I know I can’t recommend it, that’s all.

You may not be able to get a formal education, but if it’s within you capability, learning new things never hurt anybody, eh? Knowledge is power. I’m able to demand (and receive) some pretty decent compensation in my work because it’s kind of complicated and there are few that have paid the price to acquire those specialized skills. I always tell young people that you should do, for a profession, the hardest thing you brain will allow you to do. Lots of people are good with their hands, or artistic, but there will, for the foreseeable future, be a shortage of people that can solve complicated problems. If you can’t do that, try and get a skill that is specialized so that you can make a decent living (actually this is one thing that better about the European education system as compared to the US IMO).

Anyway, sorry about the rant -- it just happens that I know a bit about it and I thought I would share.


Volvorsport - 6/12/09 at 02:21 PM

the darrian , is not solely a race car chassis - its a rally car , and needs to be road registered , in fact its operating enviroment is much harder than using it on the road .

if youve designed for stiffness the strength will be there .

obviously feeding a suspension load into a flat panel is not a good idea , since it has no stiffness .

the murataya , is a grp monocoque spawned from the minari - do you suppose that might be engineered correctly ? and not fly into a million pieces .

im not being negative here .

GRP/polyester , correctly used + designed will be superior to steel in strength/weight ratio and most of the successful designs started of being built in sheds .



[Edited on 6/12/09 by Volvorsport]


boggle - 6/12/09 at 02:31 PM

i used to cut pannels for the ballistic sector.....most stuf was grp and kevlar, its extreemly strong and fairly light....i remember a company who made a composite go-kart chassis.....

so how about titanium tube as a chassis???


David Jenkins - 6/12/09 at 02:55 PM

I must admit to being amused by people who are prepared to spend a fortune to save a few Kg on what will, in all probability, be a road or track-day car. The return on investment is ridiculously low.

I once followed a long-winded and occasionally heated topic on a different forum regarding weight saving on a road-going car - the builder was spending ludicrously high prices to get the weight down on his car, against the advice of many on that forum. However, I had met that person at a show - and he must have been 18 stone, if not more! The biggest way to reduce weight on the track would have been for him to shed 6 stone... THEN he could have started to worry about reducing the weight of his car.

The steel space-frame has lasted so long because it works - it's not too heavy, it's strong, and it has quite a bit of built-in redundancy (i.e. a few things can fail before the whole structure fails). It is also easy to maintain and/or modify. I'm not sure that enough people have the skills to make a composite chassis with those qualities.

[Edited on 6/12/09 by David Jenkins]


boggle - 6/12/09 at 03:00 PM

i allways thought that if a car became too light it started to become unstable??


Theshed - 6/12/09 at 03:23 PM

That's not a rant!

Kenny you are quite right to point out the fact that most of these suggested variations on the space frame theme are vanity projects i.e. because I want to. On the other hand that's not so very different to most other hobbies.

There is a great tradition of shed built composite cars over here some good many not so good. In the main the worst ones probably never get near a racetrack or road.

What we need now is a threat entitled "Well if you really want to do that here are a few tips......."


aitch - 6/12/09 at 07:43 PM

is there a good genuinely freeware or shareware cad

id like to get started with some form of 3d design but cant afford to buy at the moment

i know ill be getting ou of date software and not the best but anything wil be better then nothing

im using vista home premium

aitch


richmars - 6/12/09 at 09:51 PM

Not free but you can get Pro E (at least in the US) for personal use for $250.
Wait a bit and I'm sure you'll soon get people showing you how to steal other CAD packages from the internet.


aitch - 6/12/09 at 09:58 PM

quote:
Originally posted by richmars
Not free but you can get Pro E (at least in the US) for personal use for $250.
Wait a bit and I'm sure you'll soon get people showing you how to steal other CAD packages from the internet.


i have "borrowed" software from time to time but always bought it if i liked it, for now im just looking for share or freeware and will buy when i can

aitch


kennyrayandersen - 6/12/09 at 10:43 PM

quote:
Originally posted by aitch
quote:
Originally posted by richmars
Not free but you can get Pro E (at least in the US) for personal use for $250.
Wait a bit and I'm sure you'll soon get people showing you how to steal other CAD packages from the internet.


i have "borrowed" software from time to time but always bought it if i liked it, for now im just looking for share or freeware and will buy when i can

aitch


try here to start:

http://www.freebyte.com/cad/cad.htm#2D3DCADSystems


kennyrayandersen - 6/12/09 at 10:48 PM

quote:
Originally posted by Volvorsport
the darrian , is not solely a race car chassis - its a rally car , and needs to be road registered , in fact its operating enviroment is much harder than using it on the road .

if youve designed for stiffness the strength will be there .

obviously feeding a suspension load into a flat panel is not a good idea , since it has no stiffness .

the murataya , is a grp monocoque spawned from the minari - do you suppose that might be engineered correctly ? and not fly into a million pieces .

im not being negative here .

GRP/polyester , correctly used + designed will be superior to steel in strength/weight ratio and most of the successful designs started of being built in sheds .



[Edited on 6/12/09 by Volvorsport]


Again, just in case you are referring to something I said, it most certainly can be done, there is no question about that. It's just not cheap, easy, or the best way to go for a personal home-brew car when you compare to a steel frame (IMO). If you got deep pockets, then nothing but time and money are preventing virtually anyone from eventually getting something to work. Knowledge will help you get there faster.


kennyrayandersen - 6/12/09 at 11:01 PM

quote:
Originally posted by boggle
i used to cut pannels for the ballistic sector.....most stuf was grp and kevlar, its extreemly strong and fairly light....i remember a company who made a composite go-kart chassis.....

so how about titanium tube as a chassis???


Titanium is a great material -- one of the few with which using a really good process you can get 100% of the parent material strength with a weld or even a casting. However, pound for pound it pretty expensive. Could it be done -- most certainly, and it would require a lot less expertise than a composite chassis. I'm not sure whether the composite (graphite epoxy) or titanium chassis would be more expensive. I'm thinking the cost somewhere around both nuts and both kidneys.

It comes mostly in round tube and due to cost you'd really like to optimize the tubes before you built. Again, this is a case where understanding structural analysis would come in mighty handy, or you could be slapping some pretty expensive parts onto your chassis unnecessarily.

You would probably be looking at commercially pure Ti rather than the 6Al4V, which is stronger, but even MORE expensive!

Ti has about 1/2 the stiffness of steel, abut more than 50% more stiffness than Al

numbers in Msi
Al = 10.4
Ti = 16
Steel = 29

As can be seen steel is pretty stiff. Fortunately the stiffness of the chassis is more dependent on the space-frame aspect rather than the individual members. Still, it's something that should be considered. I still get the feeling that people are looking a steel as the poor step child rather than the firstborn son. You can do better than steel, sure, but you will get spanked for it (some folks like rope burns -- know what I mean wink wink).


Benonymous - 7/12/09 at 12:26 AM

Agreed Kenny, steel is very hard to beat as a chassis material. I read somewhere that a welded joint can be left basically tacked (20%) and be 60% as strong as a full length weld so even if your welding skills aren't exceptional a reasonable full length weld is still very strong. The other aspect of composite vs space frame is the fact that a large amount of the chassis needs to be re-thought. A simple exchange of one material for another is not going to work. You still need to mount your engine and suspension along with a million other details, all of which need to be planned and designed.

Also, a suggestion for a basic CAD package to get started with would be Google Sketchup.

http://sketchup.google.com/

You will not be able to export a DXF file or anything like it to feed into a CNC machine or similar but it will allow you to get to grips with 3D modelling. It's free and legal.


kennyrayandersen - 7/12/09 at 02:15 AM

quote:
Originally posted by Benonymous
Agreed Kenny, steel is very hard to beat as a chassis material. I read somewhere that a welded joint can be left basically tacked (20%) and be 60% as strong as a full length weld so even if your welding skills aren't exceptional a reasonable full length weld is still very strong. The other aspect of composite vs space frame is the fact that a large amount of the chassis needs to be re-thought. A simple exchange of one material for another is not going to work. You still need to mount your engine and suspension along with a million other details, all of which need to be planned and designed.
Also, a suggestion for a basic CAD package to get started with would be Google Sketchup.
http://sketchup.google.com/
You will not be able to export a DXF file or anything like it to feed into a CNC machine or similar but it will allow you to get to grips with 3D modeling. It's free and legal.


You bring up an excellent point. If all you are trying to do is make black aluminum (quasi isotropic layup (equal number of plies in all directions (45/0/0/45)s, for a typical 8 ply layup (note the little s means symmetric about the s)), then it’s really a lot of money for not much (the density of Al is .101 Lb/cu. in. whereas graphite is about .056 Lb/cu. in. (though the home builder probably won’t get quite that light on the composite)) (Note that the cost has quadrupled or quintupled (etc.) and the time to complete has gone up by the same factor). Composites really are only useful for parts that are subjected to a maximum of two directional loading (like a shear panel). If they are loaded in all three directions (like a lug or fitting) it gets horribly difficult [since you introduce weak failure modes such as inter-laminar tension which gives only the strength of the epoxy which is only a fraction of the strength that exists in the fiber direction].

The only real way to make the composite pay off is to make the chassis completely from the beginning with the composite in mind. As I stated earlier, this would likely be some type of composite tub with a metal frame front and aft. The aft frame would tie the rear suspension together and also tie the suspension to the tub. The front frame would do the same for the engine and front suspension. Of course with a mid-engine car the rear frame would double to hold the engine and rear suspension together, and that sub assembly could be then bolted to the tub. The tub could likely be extended to gather in the non-engine end suspension with some thoughtful design.

You could probably more cost/time-effectively save weight by just making the monocoque tub out of aluminum. You’d save most of the weight for a fraction of the cost. With aluminum you would want to keep the stresses below the fatigue limit for the material. The question might be asked – is it a lowcost anymore? I guess it’s all relative. To me the concept of Locost is bang for the buck. Even the Ultima uses a steel frame… think about it.


Benonymous - 7/12/09 at 09:00 AM

quote:

Even the Ultima uses a steel frame… think about it.



So does the Norma M20!!




PHWOAAAR!


aitch - 7/12/09 at 02:10 PM

Ok here it is
i read and understand everyones posts and have little option other than to concurr that production of a composite chassis is neither the easy nor the cheap route, and defonately not the quickest form of construcion, however, i dont believe that it has to be massively expensive either, GRP in place of carbon for most of the structure will save on cost immensly, all the same forms open to car builders are also available to the homebuilt aircraft industry, steel space frames, wooden construction, rivetted aluminium and composite. without doubt steel is cheapest and quickest.

There really is an element of """ this is how i want to do it """

and for me this is still composite construction, i have looked at various ways of introducing "HARD POINTS" into the structure for mounting suspension, engine, gear box, diff etc.. and feel the easiest way is to include steel sbframes withing the GRP composite structure so that loads are introduced over a larger area, with hard points included in the honeycomb to allow forces to be distributed between inner an outer skins

Looking at other posts it would seem most peoples "low costs" build costs run upwards of £3k and up to £5k-£6k

i am looking at a build being completed in 2-3 years from now with the design phase being at least a year possibly 2 so have pleanty of time to source reasonably priced engine diff and donor compoments for the steering and suspension, all of which will be needed before the design can be finalised,

My Aim therefore is a BEC mid engined, chain drive, using modified bike wiring loom, GRP composite chassis with steel subframes (included within the composite structure) and build cost of £5000 rolling chassis (driveable) with possibly another £1000 for the body work

i have most of the tools required so there will be little expense required there and plan to out source as little as possible, if at all.

aitch

[Edited on 7/12/09 by aitch]


aitch - 7/12/09 at 08:27 PM

as im looking around trying to decide on a few other issues, mechanics and suspension im having second thoughts

for sure the engine / diff / suspension arrangement lends itself to the open arcecture that i can only get with a space frame, maybe composite tub with spaceframes frnt and rear ????? only problem here is i could end up with the worst of both worlds, complexity in design and construction and weight of steel

mmmmmmmmmmmmmmmmmmmm

aitch


kennyrayandersen - 7/12/09 at 10:35 PM

quote:
Originally posted by aitch
as im looking around trying to decide on a few other issues, mechanics and suspension im having second thoughts
for sure the engine / diff / suspension arrangement lends itself to the open arcecture that i can only get with a space frame, maybe composite tub with spaceframes frnt and rear ????? only problem here is i could end up with the worst of both worlds, complexity in design and construction and weight of steel
mmmmmmmmmmmmmmmmmmmm
aitch


Actually, you’ll find that a lot of folks do it that way. It’s not impossible to do it all composite as I previously stated, just really difficult. Using a space frame (hollow steel tubes) tied to the corners of your tub I think is the right way to go. If you search on pictures of monocoque, tub etc. you should get some good ideas. Do keep in mind that the stiffness of glass reinforced fiber is nearly 4 times less that graphite, but it is a lot cheaper (that’s why you see it used for bodies, but not so much for chassis). I know you are ‘set’ on doing it this way, so after you look at some examples of ‘working’ cars, make some sketches and post them; I, and I’m sure others, would be happy to give you some constructive comments.


aitch - 7/12/09 at 11:10 PM

quote:
Originally posted by kennyrayandersen
quote:
Originally posted by aitch
as im looking around trying to decide on a few other issues, mechanics and suspension im having second thoughts
for sure the engine / diff / suspension arrangement lends itself to the open arcecture that i can only get with a space frame, maybe composite tub with spaceframes frnt and rear ????? only problem here is i could end up with the worst of both worlds, complexity in design and construction and weight of steel
mmmmmmmmmmmmmmmmmmmm
aitch


Actually, you’ll find that a lot of folks do it that way. It’s not impossible to do it all composite as I previously stated, just really difficult. Using a space frame (hollow steel tubes) tied to the corners of your tub I think is the right way to go. If you search on pictures of monocoque, tub etc. you should get some good ideas. Do keep in mind that the stiffness of glass reinforced fiber is nearly 4 times less that graphite, but it is a lot cheaper (that’s why you see it used for bodies, but not so much for chassis). I know you are ‘set’ on doing it this way, so after you look at some examples of ‘working’ cars, make some sketches and post them; I, and I’m sure others, would be happy to give you some constructive comments.


ill probably change my mind 20 times while searching for parts, ultimately a GRP chassis although froma design point of view im thinking its a bit ambitious for a first build, although still remains my aim

aitch


Benonymous - 8/12/09 at 02:05 AM

There's no doubt you will change your mind aitch! This forum is a great place for kicking ideas around and we all do it I know I'm getting a bit single track here but I'm in love with the Norma M20 How's this for a chassis?




Triangulation in extremis! The structure is not just for torsional stiffness but also for crash safety. In addition to the multitude of tubing, the makers add shear panels to the whole structure which ends up making it look look a monocoque tub!

Just a question Kenny. If one was to construct a chassis from square steel with the intention of fitting a full set of shear panels, would you use lighter steel from the beginning?


kennyrayandersen - 8/12/09 at 08:55 AM

quote:
Originally posted by Benonymous
There's no doubt you will change your mind aitch! This forum is a great place for kicking ideas around and we all do it I know I'm getting a bit single track here but I'm in love with the Norma M20 How's this for a chassis?


Triangulation in extremis! The structure is not just for torsional stiffness but also for crash safety. In addition to the multitude of tubing, the makers add shear panels to the whole structure which ends up making it look look a monocoque tub!
Just a question Kenny. If one was to construct a chassis from square steel with the intention of fitting a full set of shear panels, would you use lighter steel from the beginning?


The mind changing for me happens when the cash has to clear edge of the wallet opening. Nothing quite defines reality like that. All the B.S., cheap talk, boasting, bravado, and Tom-foolery are stripped and you get to decide is this REALLY what I want to spend all that hard-earned cash on?

That’s a heck of a chassis there! I’d be afraid it would be too stiff in a crash, but without knowing all of the gages tube length and what not there is no way to tell. It’s certainly a bit more complex that most se7enisk chassis. I’m working on a helicopter right now, and though I’m not a crash specialist, it’s been interesting working with those guys to come up with details that actually cause the structure to fail in a predictable and sequential way.

With regard to the tube sizing: generally if you are building a tub, it amounts to a box inside another box. Since you have two walls on either side of you and you have the skin on top and bottom of that section, it would seem that you could certainly reduce the axial members since the strength comes from that bending section. The I of the upper section has gone from one 1 inch rectangular section in the typical steel frame to a whole upper skin in the case of a tub-type monocoque -- the skin providing a lot more moment of inertia, since all of that area is at the extreme fiber of the section (which decreases the stress).


twybrow - 8/12/09 at 11:12 AM

aitch - your iea of a composite tub with fromnt and rear frames sounds much more realistic. Just think about the following:

- How are you attaching composite and steel together?
- How does the load from one transfer into the other?
How are you going to jig/align the structures?

If I were you, I would start making lots of composite parts, using various materials. Get a feel for what works and what doesn't. do plenty of readings, looking on u-tube etc and build up some knowledge of the materials and processes need to use them.

If you are going to use CSM (chopped strand mat) and polyester, this is a very forgiving material to use, but very difficult to master - start practicing now! I don't think Build models to gauge stiffness/torsionsal rigidity of different designs, and cut up/test some of the composite pieces you make, to understand the effect of your skill on the performance.

I think you will get there, and I'd be happy to help where I can.


aitch - 8/12/09 at 02:10 PM

quote:
Originally posted by twybrow
aitch - your iea of a composite tub with fromnt and rear frames sounds much more realistic. Just think about the following:

- How are you attaching composite and steel together?
- How does the load from one transfer into the other?
How are you going to jig/align the structures?

If I were you, I would start making lots of composite parts, using various materials. Get a feel for what works and what doesn't. do plenty of readings, looking on u-tube etc and build up some knowledge of the materials and processes need to use them.

If you are going to use CSM (chopped strand mat) and polyester, this is a very forgiving material to use, but very difficult to master - start practicing now! I don't think Build models to gauge stiffness/torsionsal rigidity of different designs, and cut up/test some of the composite pieces you make, to understand the effect of your skill on the performance.

I think you will get there, and I'd be happy to help where I can.



i would use glass cloth and epoxy resins, slightly more expensive but will transfer stresses in a nore prediatale way with far less material and therefore weight.

the tecniques are all very well doccumented, take a look at this wing spar which demonstrates just how strong a seemingly thin walled and skinned box section can be, these spars are rated at over 9g loading http://www.cozybuilders.org/chapters/chap14.html

heres a link to the firewall
http://www.cozybuilders.org/chapters/chap15.html
and to the fusalage construction
http://www.cozybuilders.org/chapters/chap06_1.html
in order to be able to bolt on in this case the engine mount the firewall or bulhead usses plywood instead of a foam core, the forces as you can imagine in attaching a 6 cylinder rotax enfine are going to be huge and similar meathods should transfer well to a car tub

my concerns are that to interconnect steel subframes to composite tun with GRP the weight will probably end up in excess of that using one meathod of construction or the other.

i will be looking into all three meathods, all composite, all steel and combination of the two, as i plan to add a body it may be that i go for steel and gain experience biulding a composite body shell (less critical but wil gain me experience) before a second all composite build

aitch


Benonymous - 8/12/09 at 11:28 PM

Good point on the home built planes aitch. I'd forgotten how many of them are composite monocoque designs. So the problems of load paths and transitioning from tub to space frame have been solved many times over on these types of machines. As Kenny has pointed out, we can all educate ourselves and find solutions the problems we encounter with chassis engineering.

[Edited on 8/12/09 by Benonymous]


kennyrayandersen - 10/12/09 at 06:52 AM

You should be careful when looking at the aircraft website because there is more going on there than a casual reading indicates. The loads from the engine are not, in fact, dumped into the bulkhead [exactly]. There was a mention of hard points, no? What a hard point is is a point on something that is designed to carry or react a point load (like the engine support frame). Generally, the loads of an aircraft are [eventually] carried by the outer skins (fuselage bending, shear, and torsion loads). The loads from the engine must find their way to the outer skins since this is where the load wants to be (stiffest load path). So, if the engine frame attaches to the bulkhead, you would expect to see a tension fitting on the backside of the bulkhead that carries the load from the engine mounting frame through the bulkhead (hardpoint) and into the skin – this is how a ‘hard point’ functions.

The point being, is that in all of this you still don’t want to load a flat plate (or the aft wall of you tub) in a direction normal to the web (it just doesn’t work that way – there is no out-of-plane stiffness to a web (shear panel)). Instead, what you will be doing is tying the aft frame that supports the engine/suspension to the outer corners of the tub where the load then can be carried in the outer skins of your tube (and also along the axial tubes or angles that make up the corner detail of your tub. You will do this with some type of fitting.

It’s kind of all about having a place for all of the load to go. And by place, I mean a direct path.

[Edited on 10/12/09 by kennyrayandersen]


Benonymous - 11/12/09 at 05:54 AM

So is this the way it's done Kenny?

The bulkhead is blue and made of steel tube. It bolts into hard points in the tub where the red cylinders (bolts) are. This way the load is fed into the outer skin of the tub and not into the rear plate of the tub. If they were bolted into the rear plate I agree, the load on the rear plate would be trying to pull it out of the back of the tub.


kennyrayandersen - 14/12/09 at 09:17 AM

quote:
Originally posted by Benonymous
So is this the way it's done Kenny?

The bulkhead is blue and made of steel tube. It bolts into hard points in the tub where the red cylinders (bolts) are. This way the load is fed into the outer skin of the tub and not into the rear plate of the tub. If they were bolted into the rear plate I agree, the load on the rear plate would be trying to pull it out of the back of the tub.


That’s getting better. What you’ve drawn, I assume, presupposes that there is an axial member at the top of the tub. Also, the unsupported kink about mid tub will cause the upper tube to go into bending. Also, the torsional stiffness would need to be provided by the center tunnel since it won’t be supported by the sidewall.

What I was thinking a little more like is a box section on EITHER side of the tub – that being met by a bulkhead in the rear (and a corresponding one at the front of course) which has the aft frame member pretty much drawn like you have it. The actual attach point though would have to be a fitting that has fasteners in the plane of the aft bulkhead as well as a hard point (tension fitting) to the corner of the tub. I’m swamped at work or I’d try and draw at least a sketch in word or PowerPoint. Maybe in another month when my contract ends and I’m unemployed (not to worry, I put a little in the bank and there are jobs out there if you are willing to go to them (might have to do a little traveling!))!

The OP in the other chassis design thread has the aft frame narrower than the ‘tub’ (and the tub wasn’t torsionally rigid). This caused the load to be dumped in away from the corner which put the aft inboard/outboard tubes in bending. If you had a box section on either side of the tub, you could put your fitting that was tying the frame to the tub in-line with the inboard wall – you would still have a good load path and would avoid bending that way.


ceebmoj - 17/12/09 at 12:46 PM

quote:
Originally posted by aitch
i would use glass cloth and epoxy resins, slightly more expensive but will transfer stresses in a nore prediatale way with far less material and therefore weight. the tecniques are all very well doccumented, take a look at this wing spar which demonstrates just how strong a seemingly thin walled and skinned box section can be, these spars are rated at over 9g loading http://www.cozybuilders.org/chapters/chap14.html heres a link to the firewall http://www.cozybuilders.org/chapters/chap15.html and to the fusalage construction http://www.cozybuilders.org/chapters/chap06_1.html in order to be able to bolt on in this case the engine mount the firewall or bulhead usses plywood instead of a foam core, the forces as you can imagine in attaching a 6 cylinder rotax enfine are going to be huge and similar meathods should transfer well to a car tub my concerns are that to interconnect steel subframes to composite tun with GRP the weight will probably end up in excess of that using one meathod of construction or the other. i will be looking into all three meathods, all composite, all steel and combination of the two, as i plan to add a body it may be that i go for steel and gain experience biulding a composite body shell (less critical but wil gain me experience) before a second all composite build aitch


That is an interesting build diary you have linked to are there any more picks of the fire wall construction? All I can see is that the ply wood witch is covered in GRP has a metal skin applied. Another are that I would be interested in seeing in more detail if it exists is how the frame that carry the engine is joined to the tub.


kennyrayandersen - 17/12/09 at 04:58 PM

It's interesting, but the pictures are crap for actual detail. The key is that they mentioned hard points. This meaning in the analytical aerospace world means that that bulkhead would be backed up with something to give it stiffness and to provide someplace for the load to go.

Just think about it. Why would you put a punch load (out of plane no less) into a flat panel? Where is the support. That thing would be flexing big time every time you put a load on it. The biggest mistake designers make (and I'm talking about professionals with degrees), is that they just don't think about the load paths.

I'm not trying to belittle anyone by any means. What I'm suggesting is that even for trained people laying out this stuff it is not easy. For the backyard mechanic with little experience or training, it can be disastrous, or worse yet deadly.

What Beanonymous posted is certainly getting closer. Just think about if you were a loa, where would you go?


aitch - 17/12/09 at 05:09 PM

quote:
Originally posted by ceebmoj
quote:
Originally posted by aitch
i would use glass cloth and epoxy resins, slightly more expensive but will transfer stresses in a nore prediatale way with far less material and therefore weight. the tecniques are all very well doccumented, take a look at this wing spar which demonstrates just how strong a seemingly thin walled and skinned box section can be, these spars are rated at over 9g loading http://www.cozybuilders.org/chapters/chap14.html heres a link to the firewall http://www.cozybuilders.org/chapters/chap15.html and to the fusalage construction http://www.cozybuilders.org/chapters/chap06_1.html in order to be able to bolt on in this case the engine mount the firewall or bulhead usses plywood instead of a foam core, the forces as you can imagine in attaching a 6 cylinder rotax enfine are going to be huge and similar meathods should transfer well to a car tub my concerns are that to interconnect steel subframes to composite tun with GRP the weight will probably end up in excess of that using one meathod of construction or the other. i will be looking into all three meathods, all composite, all steel and combination of the two, as i plan to add a body it may be that i go for steel and gain experience biulding a composite body shell (less critical but wil gain me experience) before a second all composite build aitch


That is an interesting build diary you have linked to are there any more picks of the fire wall construction? All I can see is that the ply wood witch is covered in GRP has a metal skin applied. Another are that I would be interested in seeing in more detail if it exists is how the frame that carry the engine is joined to the tub.


thts just one of very very very many build diaries, if you do a search on cozy mk IV, Velocity, Berkut, and the list jut of this aircrafy type goes on, within the tub's there are box sections at the top edges, a central box section a bit like a transmission tunel in a car the seat back triangulates the top of the bulkhead to the bottom of the tub floor, when i get the chance ill post links to some other diaries that show some more detail..

aitch


kennyrayandersen - 17/12/09 at 11:55 PM

pics:

http://www.advancedcompositetech.net/images/engine.png

http://www.berkut13.com/engmnt.jpg

There are also several shots in the general buld area:

http://images.search.yahoo.com/images/view?back=http%3A%2F%2Fimages.search.yahoo.com%2Fsearch%2Fimages%3Fp%3Dengine%2Binstall%2Bplane%26ei%3Dutf-8%26y %3DSearch&w=400&h=341&imgurl=www.berkut13.com%2Fengmnt.jpg&rurl=http%3A%2F%2Fwww.berkut13.com%2Fberkut24.htm&size=26k&name=eng mnt+jpg&p=engine+install+plane&oid=fa55df4c80dee306&fr2=&no=18&tt=70&sigr=1143qmko5&sigi=10ris8g5n&sigb=12lpp642s

http://www.zenithair.com/design/912-firewall.gif

http://www.fly-aerotrek.com/photos/tech/912uls-engine.jpg

You’ll notice in the several pictures that the [typical] 4 mounting points are mounted outboard as far as practical on the bulkhead very near the skin. What you don’t see is any reinforcement or backup structure that might (or should) be included.

What I did find while looking for pics though was very enlightening and apropos, I ran across this Vans aircraft safety bulletin. It was interesting to say the least:
http://www.vansaircraft.com/pdf/sb08-6-1.pdf

Pages 5 and 7-10 are particularly helpful in demonstrating what I’m talking about. What we have here is a woefully inadequate design that they fix by slapping on a doubler. First, the problem occurred for the exact reasons I was mentioning; which is load was being dumped (reacted) by flat-plate bending in the bulkhead. This is a terrible load path to begin with and it really, IMO, should be backed up by a little angle fitting in the corner behind the bulkhead. A tension fitting would receive the load from the engine mount and then pass it along to the outer skins. It would weigh mere ounces and the design would be far more robust. You can fix it the way they did, but I don’t think they are making it a particularly robust design that way and they are likely to have problems with it in the future. I’ve always fancied the idea of building a plane and generally Vans has a decent reputation, but after seeing these photos, I’d take a HARD look at the one I wanted to build before making that leap (unfortunately not many buyers are going to have the technical expertise to evaluate the structural integrity of what they are buying and will rely on the manufacturer to do that for them…).

The whole problem would have been avoided in the first place had they properly backed up the attachment (i.e. their ‘hard-point’ was actually a hard point!) like I was yammering on about in the earlier posts. This also demonstrates what I was saying about even professionals (at at least people professing to be such) fairly frequently have trouble executing.

Now, let’s get back to cars. The fore and aft sub-frames that the OP is talking about mounting to a tub are going to be quite similar to these and if those points are not backed up with a little structure, you can expect the same [crappy] results. IMO, there is no problem to do it; it just has to be done with a little forethought and common sense. You can do it by heavily reinforcing the bulkhead, or by making a light-weight tension fitting, but you must do something. These days you can get on-line, upload your solid model and have someone machine it for you relatively inexpensively. There is not a good reason to do a half-assed job of it just because you don’t have the full shop (you should note though that lacking the full shop you would have to have the FULL budget).

I know I only mostly work on fighter aircraft and helicopters, so I have some learning to do about cars. Nevertheless, the principles are universal and if understood and applied properly it will result in a safe and robust design. In the end, we are back to where we started eh? Stick with steel unless you want to play the Man of La Mancha.


kennyrayandersen - 17/12/09 at 11:57 PM

Now here is a nice looking engine attach:

http://www.flycorvair.com/jan04601h.jpg

right out to the edge (where the skin can get the load and where the bulkhead won't get the crap knocked out of it. Also note the overlapping stiffeners make a natural spot for the load to be dumped into (guess I'll have to take another look at that Zenair).

[Edited on 18/12/09 by kennyrayandersen]

[Edited on 18/12/09 by kennyrayandersen]


kennyrayandersen - 18/12/09 at 10:48 PM

The following is a response to someone looking to build a monocoque-type chassis in another yahoo group I frequent. The responder lists some sources that would be applicable to trying to do a composite tub (or other monocoque construction). There is probably some good information there as the sources sound like they are relatively authoritative. I you , or anybody else wants to do this type of construction (composite or not), they should, at a minimum, do a a little 'light' reading:


Woody,
Earlier this month you posted you were looking into producing trikes and wanted
advice on designing monocoques.
I was browsing in a bookshop last week and came across a book "How to build a
motorcycle engined racecar" by Tony Pashley, published in 2008 by Veloce in the
UK. The author built several hillclimb racers using tubing and then alloy
honeycomb sheets. The builds were covered by articles in Racecar Engineering
magazine. The book incorporates these articles and expands on that info. I had a
quick skim and it had some useful info that might help you. There are several
other books you may have or know about but I will mention them here for other
members as well who may be thinking on the same lines.
"Competition car composites: A practical guide" by Bryan O'Rourke and "The race
car chassis HP1540" by Forbes Aird. I have his earlier work called Race Car
Chassis - Design and Construction (1997) in my library and found it to be very
good. Composites can make a lighter, stiffer and stronger chassis than a
comparable steel tube chassis. The downside is cost and practicality. They are
difficult to make - e.g.drilling accurately in a honeycomb where the drill runs
off and even harder to repair. They also require a lot more design skill and
knowledge than a simple tube frame. But we need builders who can push the
envelope so best of luck if you follow this route. Have you considered using
timber plywood and epoxy resin. This is very suitable for one-off or small
production runs. The Marcos sportscar designed by Frank Costin in 1960 and the
Mosquito bomber in WW11 were built using this method. Costin used Aerolite 308
two pack resin glue which is still sold today as Selleys High Strength glue. But
there have been great advances in epoxy resins and glues. I have been working on
a timber monocoque and am using an aerospace glue called Hysol. It is used to
stick the skins to airliners. Mind you at >$100/litre it is not cheap. If you
want to study Costin's designs you could look for a copy of "Flying on Four
Wheels" by Denis Ortenberger. This is an excellent study of Costin and his work.
You might also look at the West epoxy system used in boatbuilding. The epoxy is
called Bote Cote and is used with plywood or balsa to create amazingly strong
composites.
I know a few members are looking at making several trikes. If you are, a 'must
read' book IMHO is one by Walter Korff called "Designing Tomorow's cars: from
concept, step by step, to detailed design" Published by M-C Publications in
1980. (there are 6 copies on Amazon). I found a copy in our State Library
several years ago and photocopied it there as you can not borrow it. He is/was
an excellent engineer who consulted to the large auto firms like Chrysler and
also built a range of small cars including trikes. Robert Q Riley's efforts
remind me of Korff's work. Many of his design views appear in later works by
other authors. E.g. Korff is the oldest reference I have found to the 'circle of
traction' concept in suspension design. More importantly, he has a whole section
in the book on what you have to do to move from a concept car like a running
prototype - essentially what we are trying to build - to production. It is
daunting reading !!
I shall suggest to our Moderator Joe that we create a page where members can
post book titles they have found helpful. This could act as a resource and
greatly assist our many new members, indeed all our members, to complete their
builds.

Cheers

Grant Campbell


ceebmoj - 22/12/09 at 08:34 AM

witch forum is the post from?


kennyrayandersen - 22/12/09 at 09:22 AM

http://autos.groups.yahoo.com/group/3WheelerReplicas/message/4745

which is a bit odd actually as most of the guys are doing something rather Morgan like, though one guy did a Messerschmitt Record Breaker


Sam_68 - 9/1/10 at 12:48 PM

Sorry - late to the party, I know: I don't visit the Locost forums often, these days, so I've only just come across this thread.

quote:
Originally posted by djtom
Westfield made a composite chassis car a few years ago. FW400 I think it was called. 400kg but OMG expensive.


I own one of the three FW400's left in this country. The chassis is a very nice piece of work, designed by an ex-Lotus F1 Chief Designer by the name of Martin Ogilvie and manufactured by ex-Lotus composite guys. The cars that were built (there were only 6) were 'pre-production' models and as such the monocoque is made up of lots of flat panels and simple mouldings bonded together. If they'd continued production, the intention was to simplify it into fewer mouldings.

As such it's pretty much a half-way house between a 'proper' moulded and autoclaved tub and the sort of routed-and-folded flat-panel tub being discussed above.

If anyone wants further details or photographs (or to take a look at the car), give me a shout.

With regards to the routed-and-folded approach:

quote:
Originally posted by StevieB
A monocoque se7en using this technique should be fairly simple (famous last words...)...


Martin Ogilvie pioneered the bonded honeycomb panel approach with the Lotus 87 F1 car.

He built an electric 'Westfield' Seven called the Wisper (Westfield In Structural Plastics Electric Roadster) that uses the routed-and-folded approach. He was kind enough to talk me through the process in some detail when I took the FW over to see him a couple of years ago, and (the way he tells it - and bearing in mind he's a talented and experienced F1 designer, not an amateur in a shed), it really is as simple and straightforward as it sounds. He reckoned that once you've got the design, the actual routing, folding and bonding of the main tub structure was basically a day's work.

There's an article on the car >>here<<, which includes a couple of photographs where the chassis tub can be seen, but unfortunately it majors on the elecvtric power rather than the chassis construction.

As others have already identified on this thread, the bit that needs more thought and care is the feeding of loads into the tub. The FW400 (and, I think) the Wisper use clever threaded 'top hat' bobbins that screw into each other from either side of the panel, being bonded in place as they do so. You'd need quite a few of these machining up, which would add to the cost of the tub, but it's a simple and elegant solution.... provided you have enough structural knowledge to place them correctly.


Paul_C - 11/1/10 at 12:29 PM

An interesting read is 'Motor Vehicle Structures: Concepts and Fundamentals' by Jason C. Brown, A. John Robertson and Stan T. Serpento, Butterworth Heinemann. It helps in identifying loads and load paths in a monocoque.

I have used Google SketchUp to draw up my guess of the Costin Amigo structure and produced a 1/10 th scale cardboard model which helps to visualise the load paths.
I have also perfomed a crude analysis of the torsional stiffness using the LISA-Finite Element Technologies software, available at www.lisa-fet.com.

From what I saw on a Plane is Born on the telly where they appeared to bolt directly through a tube with no re-inforcement to prevent crushing. And also where the undercarriage didn't have any proper suspension it's probably best not to use Kit Planes as best design practice.


ceebmoj - 15/1/10 at 09:11 AM

link that some one else has posted up of the design of a formula student car. witch has a hunycome chassis

http://www.locostbuilders.co.uk/viewthread.php?tid=128266&page=1#pid1077629


SPYDER - 25/2/10 at 07:41 PM

aitch you have U2U.
Geoff.