Jermyn
|
| posted on 22/9/04 at 04:00 PM |
|
|
Chassis Analysis: A Waste of Time?
OK, I just want to start of this debate stating that this topic is not intended to p*ss anyone off. 
I was appoaching my build thinking that I should draw up my own chassis in CAD and do multiple tests to determine the strengths and weaknesses of a
book chassis and how it could be improved upon. I also have noticed a number of other builders out there doing the exact same thing, and in some
cases, taking it to extremes.
After some time reading others websites, papers, dissertations and studies, I came up with these questions:
Haven't the mods made by Mcsorley, Cymtrix and others who have already studied the Locost chassis dynamics had the greatest impact?
Would the limited mods I or others make at this point in time will be superfluous to the needs of any average 7 builder? So why bother?
Certainly improvements can always be made to something and I'm not trying to quash anyones creativity. It's just that I am wondering why I
put in so much effort for negligible benefit on a frame design that is tried and tested?
I could understand it for those building middies or straying away from the book chassis by a great deal. But, on my first build, it's going to
be fairly close to the book.
Comments?
If life is a race do you want to be the first one to finish?
|
|
|
|
|
andyharding
|
| posted on 22/9/04 at 04:07 PM |
|
|
If it ain't broke don't fix it.
Are you a Mac user or a retard?
|
|
|
Mix
|
| posted on 22/9/04 at 04:43 PM |
|
|
As far as my build goes I have enjoyed the challenge of using the book as a guide and adapting the basic design to what I want.
Building one of these cars from scratch is always going to require an individual approach, everybody will have their own views as to the best way of
achieving the aim. The contributions of others on here are invaluable but at the end of the day the car you build will be unique and as such will not
conform to any set standard. We as builders decide on the level of innovation/development we incorporate into our cars and by our mistakes/sucesses
may assist others in the future.
I suppose what I'm trying to say is to do your own thing but don't feel it's all been done already.
Mick
|
|
|
derf
|
| posted on 22/9/04 at 05:56 PM |
|
|
I actually enjoy looking at the different ways that everyone else has come up with to improve chassis strength. I seem to remember one car that came
up with an engine cradle and increase engine bay strength by loke 1000 times. I am waiting (or rather I waited and now I've built my chassis so
it's too late) to see someone come up with an ingenious design that is radically different, but keeps the same external dimension. And an extra
mind thrown in could be the one that has that radical idea
|
|
|
Cita
|
| posted on 22/9/04 at 06:01 PM |
|
|
Playing with autocad will cost you nothing except some time.
If you have some idea's go for it when you still can use your free time for drawing up some frame structures,once you're in the
construction mode it's difficult to get back into the designer mode if you know what i mean
|
|
|
Alan B
|
| posted on 22/9/04 at 06:54 PM |
|
|
Couple of points.
Although Cymtriks mods seem like they make sense, they are untested as far as I know.....someone please report back when tested.
The McSorley plans aren't structural improvement mods, simply a redraw and correction of the base plans and some plus variations.....I'm a
friend of Jim's and I'm sure he'll agree with me. Don't get me wrong, Jim did a LOT of work and his plans are great, but
I'm sure he claims no structural changes.
|
|
|
Cita
|
| posted on 22/9/04 at 07:31 PM |
|
|
There seems to be some doubt about Cymtrics proposed modofications.
I think we all agree that they dont make the chassis any weaker.
So if they are not weakening the chassis,make it lighter and probably stiffer why the hestitation to use his modifications?
|
|
|
Alan B
|
| posted on 22/9/04 at 08:15 PM |
|
|
My only doubts is they aren't tested....I'm fairly sure they will be OK....
My gripe is people treating his mods like gospel without some feedback.
No offence to Cymtriks I'm certain he his great at his job, but his job is not designing car chassis...again someone correct me if I'm
wrong...
All I'm saying is don't carried carried away, just treat them as theorectical until tested..
|
|
|
Mark Allanson
|
| posted on 22/9/04 at 08:45 PM |
|
|
Weakest link engineering never seems to get mentioned here. If you are perpetually strengthening and stiffening parts (including the chassis) at some
point something has to give. If the loads are evenly spaced over the suspension joints, brackets, dampers, and chassis, there will be no appreciable
problem, but if the chassis in infinately stiff, all the stress will be on the suspension and the WELDS. Its the old unstoppable force meeting the
immovable object syndrome.
I am not saying that a stiff chassis is a bad thing, but taking things to extreems will cause unthought of problems
If you can keep you head, whilst all others around you are losing theirs, you are not fully aware of the situation
|
|
|
stephen_gusterson
|
| posted on 22/9/04 at 09:26 PM |
|
|
merc claim that the smart car is rigid, and 'exploits the crumple zones of the other car'.
Its strong in a sideways crash as the wheels are closely spaced and impact will be taken at those points on the side.
yeah right.
marketing bo&^cks.
how do you exploit the 'crumple zone' of a 4x4 or a volvo!
atb
steve
|
|
|
Mark Allanson
|
| posted on 22/9/04 at 09:58 PM |
|
|
The A series mercs have a twin level floor so are quite good in side impacts, but I agree about the marketing blurb - merc's aint wot they used
to be.
4x4's as a rule are death traps, if you have a med to heavy crash, coupled with the false sense of security they give, you are in trouble.
Volvos. They are safe, they have VERY soft extremities, and absorb shock very well, but the lethargic handling, they are more likely to need it!
If you can keep you head, whilst all others around you are losing theirs, you are not fully aware of the situation
|
|
|
TheGecko
|
| posted on 22/9/04 at 11:40 PM |
|
|
Jermyn,
I think you make a quite valid point. Many (many) cars have been built to the book design or close to it. So they fall apart after the first 500
kilometers? Generally, no (need to make the distinction between design quality and build quality  ).
If you're looking for a stiffer chassis, try the Australian chassis mods. We have to actually pass physical tests of the chassis here so the
Aussie mods are tested and known to work.
I'm building a mid-engined car only loosely inspired by the Locost so I'm doing a fair bit of planning and checking to ensure that my
chassis is stiff and strong enough to: be legal, handle well, and be safe.
If you're building a book Locost in the UK, I'd say leave it alone.
Edit: just noticed you're in the US - the only reason you might make significant chassis mods is to deal with different donor components.
Dominic
[Edited on 22/9/2004 by TheGecko]
|
|
|
Cita
|
| posted on 23/9/04 at 02:27 AM |
|
|
Heard this before also but what exactly are the Australian modifications?
How are these cars tested on tortional stiffnes?
Just want to learn
|
|
|
Hugh Jarce
|
| posted on 23/9/04 at 04:18 AM |
|
|
quote:
Originally posted by Cita
There seems to be some doubt about Cymtrics proposed modofications.
I think we all agree that they dont make the chassis any weaker.
Who says they don't? Thet're not tested! (I'm sure they won't actually make things worse though).
Check out some of the Aussie sites, they go into great detail about ADR testing.
The pay isn't very good , but the work's hard.
|
|
|
kiwirex
|
| posted on 23/9/04 at 08:59 AM |
|
|
I started doing an analysis of my chassis using a 'simple' method that I'd read in a book.
the theory of the method was you estimated the down forces at the key joins in your chassis, and used calculation of moments to make sure you'd
got everything right.
Then you drew the forces on a drawing board - in the direction of your tubes, but the length proportional to the force on the tube.
If your intersection has two unknown forces on it then you can estimate the forces on the remaining tubes by using lines parallel to the tubes and
measuring the result.
My explanation ain't that good, but there's a (sort of point).
After getting halfway down the chassis and finding a bit with more than two unknowns (necessitating a guess), I started from the other end. I got 4
tubes in and discovered that I could estimate the stress on the same tube from two different intersections, and they were different by a factor of 10.
Which kind of put me off the methodology somewhat as it hadn't mentioned anything like that.
Once I flagged it away, I decided that, while it's probably not worth doing the full FEA on the chassis (for me, anyway), looking at other folks
chassis ideas, and *understanding* them, then pinching what makes sense, seems like a good idea.
So I'm going to do mostly cymtriks mods, with McSorley's revised front end (not for strength, for manufacturability).
A point? Oh yeah, um.
Worth thinking about the chassis design before you cut tubes,
But only you can decide if it's worth plugging the whole thing into an FEA package or whatever.
Cheers,
Greg H
|
|
|
MikeRJ
|
| posted on 23/9/04 at 10:23 AM |
|
|
quote:
Originally posted by Hugh Jarce
Who says they don't? Thet're not tested! (I'm sure they won't actually make things worse though).
The modifications that triangulate the front part of the chassis, are pretty obviously going to make the structure stiffer, at least in that area.
FEA is a very usefull tool, but with a little bit of thought it usualy possible to see where the forces in a chassis are going to be distributed and
where the bigest improvements are likely.
|
|
|
Terrapin_racing
|
| posted on 23/9/04 at 11:10 AM |
|
|
I'd go with Alan on this. Don't treat any mods as gospel. We can all come up with potential improvements - However, all these either based
on experience or theory. This is why manufacturers crash test cars (any volunteers for doing the locost ones?). To put things in perspective - I know
of a single seater driver who made some structural modifications to the front of a car that greatly improved the stiffness and handling over the
original design - Great!
..... until he had a crash and ended up with one of his additional steel tubes impaling his leg in a frontal crash!
Sometimes there's a trade-off
|
|
|
pbura
|
| posted on 23/9/04 at 01:13 PM |
|
|
quote:
Originally posted by Cita
Heard this before also but what exactly are the Australian modifications?
Information about all the chassis mods commonly discussed (Aussie, Cymtriks, Linton, DSK) is available here:
http://locost7.info/mirror/chassis.php
My own $0.02: Just as adding more tubes to a chassis results in diminishing returns after a while, a fresh FEA analysis is unlikely to yield much in
further improvements in chassis rigidity. If you want to do it, there's certainly no harm, but there may be more of a payback in trying a
design with larger section tube, or a ladder/space frame combo, or a box-section chassis.
Most of the proposed mods appeal to common sense and don't add a lot of weight, so why not use them? I'm grateful for the work
that's been done, and plan to use several of these ideas. Point taken on the impaling, however 
Re: stiffness to the point of danger, the greatest stiffener would be a roll cage, which are frequently used without chassis failure.
Pete
|
|
|
Rob Lane
|
| posted on 23/9/04 at 01:41 PM |
|
|
The Locost design was copied from the Westfield pre-lit which in turn was an interpretation of a Lotus 7/Caterham chassis. (Strange how each sued each
other
Westfield, in order to get away from the anti-roll bar cum suspension arm a la Caterham, developed the wishbone with 'coffin' front
end.
Now as far as I'm aware they did no CAD analysis of the design but applied engineering priciples and suck it and see.
Over the years I watched threads as newcomers come and go, each deciding that the chassis needed CAD analysis in order to improve it.
As far as I can tell, the only real mods without penalty are the minor additions that the Oz guys do. I've incorporated them in my chassis and
the weight difference was something like 8kg.
However the measured benefits as done by the Oz testers bears out that these simple mods stiffen the chassis by approx 50%. The results are somewhere
in Yahoo Locost list files.
As has been stated, you're wasting nothing but time when doing an analysis. If it makes you feel better , do it.
Incidently, after the Oz mods there was a distinct improvement to the cars handling along with a loss of chassis 'creaks'. Mostly from
the rear panel where the rivets became 'loose' with the chassis flex.
Actually they were still tight when examined but moved under strain.
[Edited on 23/9/04 by Rob Lane]
|
|
|
Rob Lane
|
| posted on 23/9/04 at 02:02 PM |
|
|
I've just looked at the mirror site chassis jpg of Oz mods. These are not the mods I was supplied with a couple of years back by Tricky.
The mods are more like the frame mod.jpg but some items missing such as gussets at rear.
|
|
|
Cita
|
| posted on 23/9/04 at 05:01 PM |
|
|
Best thing is to leave everything as per the book?
All those comments are very confusing for a not too clever guy like me.
I dont think Cymtric is sucking his knowledge out of his thumb(as our politicians do!)and his modifications are not uterly complex or expencive so why
not take them and use them.
I dont want to step on anyone's toe here but i got the feeling that "i did'nt come up with this "improvement" so it
probably is no good" is making a too big a fuzz about this.
All IMHO of course.
|
|
|
stressy
|
| posted on 23/9/04 at 07:52 PM |
|
|
I completed an FEA for a book locost chassis around 5 or 6 years ago. My reason for this was to try and get to grips with the constraints approach in
order to analyse somebody elses design for them.
In the process of doing this i made a number of mods with regards to impact safety and geometry changes. My own car incorperates these changes.
They are related to the effective strength of the floor, the triangulation and geometry of the transmission tunnel and forward floorplan (i.e. my
engine just clears the footwell bulkhead), rear and side reinforcement, and front end lateral triangulation.
My advise to anybody starting out is to go with the book design or thereabouts, maybe apply the common basic mods.
If your looking to design a considerably improved car then starting with the standard book chassis is NOT the place. In fact starting with the
chassis structure is not the place.
With regards to cymtriks and other detailed mods: the panelling of the transission tunnel and steel floor is not a bad idea as it gives greater
strength in the passenger region, the y shapes triangulation is ala caterham so not likely to be too far wrong, but i would personally be rather
hesitatant as to the triangulation between the wisbones, diagonal fore and aft. In a front ender you may loose the crushable section of the nose, this
loads up the next set of tubes back, one of which ends up (on a book chassis) between your toes, putting large bending stresses in the frame.
Many 7 derivatives use this triangulation at the front but utilise lighter gauge material from the foot well forward in order to leave some sutable
failure points for crash tolerance.
Only my opinion for what its worth.
WHO DARES SPINS
|
|
|
Hugh Jarce
|
| posted on 23/9/04 at 10:08 PM |
|
|
quote:
Originally posted by MikeRJ
quote:
Originally posted by Hugh Jarce
Who says they don't? Thet're not tested! (I'm sure they won't actually make things worse though).
The modifications that triangulate the front part of the chassis, are pretty obviously going to make the structure stiffer, at least in that area.
I fully accept some of Cymtrics' suggestions do make sense, but my comment was to emphasize that no amount of theory, no matter how
logical at face value, should be trusted until tested and proven one way or the other.
No offence to Cymtrics, but he doesn't appear to have finished a Locost let alone put his theories to the test, yet he seems to have
attained some (probably wholly unintentional) elevated status as a FEA guru.
To Cita: read what you like on here, but bear in mind, that unless someone states a certain setup worked for them over x miles of consistent road or
track use, then it's all just guff in the wind, or at best, educated guesstimates.
Don't build your car and put your life at risk just because someone posts alot of unqualified or untested thoughts on a certain subject.
If you can't sort through the theories, or don't want to be a guinea pig (I'm all for advancement), just follow the hundreds who
have built the version that works.
The pay isn't very good , but the work's hard.
|
|
|
Jermyn
|
| posted on 23/9/04 at 10:41 PM |
|
|
My initial inclination was designing for performance. I hadn't even given any consideration to safety. But, now that so many of you have brought
it up, I think you have touched on a truly important aspect of this topic. I don't want to add mods to my car that could compromise my safety in
the event of a crash. And, being that I don't have a crash test course set up in my basement, I would have no way to check if my new mod
wouldn't end up in my chest in a crash. OUCH. 
When I started this topic, I was going to use some of the Mods already out there. Now, I'm reconsidering strenthening mods altogether. Certianly
stiffening one point in the chassis has the propensity to defer force to other points, possibly causing those to fail under stresses they
weren't initially designed to handle. (I still might consider the Aussie ones, since they seem to have safety in mind first and performance
second. anyone got a link they can post?)
ROB- do you still have the old Oz mods you could post?
Plus, I think the issue Thegecko brought up with design quality v/s build quality is also relevant. The welds on the frame have every bit as much to
do with the stiffness as any other part. So, am I to believe internet posted FEA results of some yo-yo who had no business even holding a welder?
I guess I enjoy reading their results regardless. Especially since they come up with these mythological strength increase numbers.
I think I'm still going to draw it up in cad just to make adjustments for my RX7 donor car rear end. But, as far as strength testing
goes....i'll probably leave it alone
If life is a race do you want to be the first one to finish?
|
|
|
Rob Lane
|
| posted on 24/9/04 at 07:41 AM |
|
|
Jermyn,
The Oz mods are based upon physical findings rather than software analysis.
The chassis is put through tests by their 'rules' engineer. See the posts on the mirror site linked in above post.
I've tried to find the actual measured results but I've probably deleted them.
I will see if I can find the post from Oz with the simple mods. Certainly the first and most easiest is the two braces in front frame, turns it into
a 'W'
|
|
|
