Hi all
Bit of background first.
Car details
weight 600kg
split 40F 60R
Kpi 0 deg
Scrub radius 26.65mm
King pin Offset 0mm
tire width 215mm
Front tire radius 313mm
Now then i am in the process of deciding how much caster to incorporate and this will determine the trail since im not using any king pin offset or
(pin lead)
Obviously the tire radius is quite large so i do not need much caster to create the trail, and dont want to much caster to prevent weather cocking as
someone kindly pointed out to me in another post. however i need some self centering, essentially enough to pass IVA
so the question is how much mechanical trail??
cheers tilly
have you read this book?
click
i still reckon you need KPI! kpi helps lift the side of the car when you turn the wheel (ie lifts the right side when you turn right) and the weight
helps self centering. well, kpi and castor, but i don't think castor itself will. reckon you're onto plums without kpi, thats all i'm
saying on it.
i understand you comment about kpi however the way i see it, kpi's conns outweigh its pros's, i dont want this to come across wrong as it is
very easy to sound rude when typing but that is not my intention i assure you. i would like to point out that the lifting of the car due to its KPI is
not a desirable characteristic.
the trail should be able to provide adequate self centering, think of a caster wheel on a trolly it has no KPI and no caster however it uses a King
pin Lead to create mechanical trail which allows it to self center, i now this is a rather extreem example but it shows the concept.
cheers tilly
Castor & KPI are like Morcambe & Wise
KPI is the best source of steering self-centering.
The trail will create weather cocking even if the actual castor angle is low, how much will depend on the moment of th lateral force on the tyre
contact about the virtual king pin.
Not meaning to sound rude but you will need some KPI, there is a reason that virtually every single road and race car ever designed has it.....
Also in general Mechanical trail is usually left as a product of the caster angle and wheel size, there are very few cars were the wheel hub centre is
not on the centreline (between the pivots) of the upright.
One thing (amongst many others) that I have learnt when at uni' studying motorsport design and out in the real world designing race cars is that
there are many things that seem like a good idea on paper or after reading a book (no KPI, Roll centres under ground, solid axles) but in real life
they make horrible handling cars that don't win races or get sold to the public.
quote:
Originally posted by Doctor Derek Doctors
you will need some KPI, there is a reason that virtually every single road and race car ever designed has it.....
Also in general Mechanical trail is usually left as a product of the caster angle and wheel size, there are very few cars were the wheel hub centre is not on the centreline (between the pivots) of the upright.
DDD
Sorry to pick on you but you were the last contributor! If everything remains the same we will never progress. I guess that's OK if you only want
to be one of the crowd but it does stifle things a bit don't you think?
Sure a lot of progressive ideas have turned out to be poo but by analysing the reasons we progress just as much as the ideas that work straight out of
the box.
So why is no kpi such a bad idea? Come to that why is "the wheel hub centre not on the centreline (between the pivots) of the upright" a bad
idea. Seems to me that Tilly is balancing pro's and cons to come up with a design and is being open about it. If there is a logical engineering
reason why not I think we have a duty to spell it out!
For my part I don't know of one apart from it's jolly difficult to package all the stuff into the wheel and get zero kpi - I know I've
tried and failed! I've seen Tilleys drawings and it seems to be possible due to a clever twist. But what is this overarching reason not to have
zero kpi?
Cheers!
Check out this link to a forum with quite a respected contributor on the subject
http://www.eng-tips.com/viewthread.cfm?qid=113181&page=8
Cheers!
KPI causing the front of the car to rise when the steering is turned a bit of a red herring because (ignoring effects of wheel width) castor
causes the nose of the to fall as the steering is turned.
The relationship between camber castor and kpi and the weight/download carried by the front wheels needs careful consideration. A major factor
involved that has to be considered are the tyres.
Using the Cortina Mk3+ as an example the tyres were much heavier loaded than on any car we are likely to design so nearly all of the steering
self-centering and stability came from the self-aligning torque generated by the deflection caused in the radial tyres.
Getting a near zero scrub radius with wide wheels and a low KPI is very difficult because of wishbone clearance and steering lock considerations.
Interestingly , production cars (Audi) with strut suspension can actually have negative scrub. Back in in the 1970s and 1980s racing saloon cars
were built with a ridiculous ammounts of scrub radius because racers were using very wide wheels on suspensions (for example Capri) that were
designed before tyre widths increased.
An increased scrub radius is not desirable but not a major problem in tarmac use.
Incidentally the only cars I can think of that had the stub axle axis offset from the king pin centreline to give mechanical trail were the Hillman
Imp and the rear engined Fiats ----- both types very prone to weather cocking.
[Edited on 14/4/11 by britishtrident]
Right i was hoping that this wasnt going to turn into a you need KPI you dont need KPI but it has.
This discution has happen a couple of times already and since knowone has been able to give a definitive "you need KPI because" or
"Zero KPI is a definate No NO because " then i see no reason why my choise to have no king pin inclination should be recived so badly..
I have given the basic geometery in my first post and what i was hoping for was some responces to establish some aproxermations of a suitable caster
angle and resulting trail, now i have made a staring estimate of 5.5deg caster which results in 30mm mechanical trail. this seems reasonable to me
since i am awere of someone who has build a similar weight car the the same aproximate weight distribution with about 25mm of trail and there steering
is very light and has some self centering
thanks
tilly
quote:
Originally posted by v8kid
DDD
Sorry to pick on you but you were the last contributor! If everything remains the same we will never progress. I guess that's OK if you only want to be one of the crowd but it does stifle things a bit don't you think?
Sure a lot of progressive ideas have turned out to be poo but by analysing the reasons we progress just as much as the ideas that work straight out of the box.
So why is no kpi such a bad idea? Come to that why is "the wheel hub centre not on the centreline (between the pivots) of the upright" a bad idea. Seems to me that Tilly is balancing pro's and cons to come up with a design and is being open about it. If there is a logical engineering reason why not I think we have a duty to spell it out!
For my part I don't know of one apart from it's jolly difficult to package all the stuff into the wheel and get zero kpi - I know I've tried and failed! I've seen Tilleys drawings and it seems to be possible due to a clever twist. But what is this overarching reason not to have zero kpi?
Cheers!
Interesting I thought the Hillman Imp has front swinging axles with a true king pin - basically a beam axle cut in half. Blessed if I can see how they
get any spindle offset with this setup you don't have any pics do you I'm very interested?
Also can't find any reference to Imps "weather cocking" what do you mean by it - as I remember the handling was OK unless you got a bit
overenthusiastic!
Cheers1
quote:
Originally posted by v8kid
Also can't find any reference to Imps "weather cocking" what do you mean by it - as I remember the handling was OK unless you got a bit overenthusiastic!
Cheers1
quote:
Originally posted by Doctor Derek Doctors
quote:
Originally posted by v8kid
DDD
Sorry to pick on you but you were the last contributor! If everything remains the same we will never progress. I guess that's OK if you only want to be one of the crowd but it does stifle things a bit don't you think?
Sure a lot of progressive ideas have turned out to be poo but by analysing the reasons we progress just as much as the ideas that work straight out of the box.
So why is no kpi such a bad idea? Come to that why is "the wheel hub centre not on the centreline (between the pivots) of the upright" a bad idea. Seems to me that Tilly is balancing pro's and cons to come up with a design and is being open about it. If there is a logical engineering reason why not I think we have a duty to spell it out!
For my part I don't know of one apart from it's jolly difficult to package all the stuff into the wheel and get zero kpi - I know I've tried and failed! I've seen Tilleys drawings and it seems to be possible due to a clever twist. But what is this overarching reason not to have zero kpi?
Cheers!
Because it just doesn't work. I know that doesn't sound right but 50+ years of motorsport developement have always led to the same conclusion. I could spell it all out here but there are plenty of massive books on the subject as well as 24 F1 cars that demonstrate it better than I can in my short lunch brake....
....but think about this, a shopping trolley self centres the wheels as it goes forwards (from the froward motion), but there is no resistance to side loading except from the steering and as you increase the mechanical trail the lever ratio of this load is increased meaning it is harder to counteract this load. This load in corners will come from the centripetal acceleration (centrifugal force to the lamen) of the car rotating around the corner, thus the more mechanical trail you have the harder is will be to steer around a corner and it won't just be trying to self centre it will be trying to follow the load path thus steering in the opposite direction! It will also be a nightmare to resist side loading from side wind and bumps. With with KPI on the other hand the it self centres to the chassis via gravity and the centripetal acceleration acts directly on the centre of hub so it is not trying to steer the car, the increased resistance you feel is from the increased load on the outside wheel working harder to centre the wheel because of KPI.
To try and sum up in a sentence Mechanical trail self centres to the direction of the load acting on it (cornering force, wind, bumps etc) but KPI self centres to the cars chassis so it will always try to go straight, whereas a shopping trolley will try to go where other outside factors are pushing them
When I was a wide eyed design student we read the books and weighed up the pro's and cons and decided that we knew better and had new ideas.... and for every idea there was always an example of why it didn't work. For instance placing the IRC under the ground seems like a great way to counteract the CoG and stop the car leaning in a corner, in theory it works, in reality it has been tried to death and it always causes odd handling characteristics.
As for lack of progress? I'm sure Adrian Newey or Ross Brawn would love to be told how they have made no progress.... oh wait.
quote:
Originally posted by tilly819
quote:
Originally posted by v8kid
Also can't find any reference to Imps "weather cocking" what do you mean by it - as I remember the handling was OK unless you got a bit overenthusiastic!
Cheers1
Weather cocking is a tendency to turn into wind when exposed to a sidewind /crosswind
quote:
Originally posted by tilly819
This discution has happen a couple of times already and since knowone has been able to give a definitive "you need KPI because" or "Zero KPI is a definate No NO because " then i see no reason why my choise to have no king pin inclination should be recived so badly..
quote:
Originally posted by tilly819
quote:
Originally posted by Doctor Derek Doctors
quote:
Originally posted by v8kid
DDD
Sorry to pick on you but you were the last contributor! If everything remains the same we will never progress. I guess that's OK if you only want to be one of the crowd but it does stifle things a bit don't you think?
Sure a lot of progressive ideas have turned out to be poo but by analysing the reasons we progress just as much as the ideas that work straight out of the box.
So why is no kpi such a bad idea? Come to that why is "the wheel hub centre not on the centreline (between the pivots) of the upright" a bad idea. Seems to me that Tilly is balancing pro's and cons to come up with a design and is being open about it. If there is a logical engineering reason why not I think we have a duty to spell it out!
For my part I don't know of one apart from it's jolly difficult to package all the stuff into the wheel and get zero kpi - I know I've tried and failed! I've seen Tilleys drawings and it seems to be possible due to a clever twist. But what is this overarching reason not to have zero kpi?
Cheers!
Because it just doesn't work. I know that doesn't sound right but 50+ years of motorsport developement have always led to the same conclusion. I could spell it all out here but there are plenty of massive books on the subject as well as 24 F1 cars that demonstrate it better than I can in my short lunch brake....
....but think about this, a shopping trolley self centres the wheels as it goes forwards (from the froward motion), but there is no resistance to side loading except from the steering and as you increase the mechanical trail the lever ratio of this load is increased meaning it is harder to counteract this load. This load in corners will come from the centripetal acceleration (centrifugal force to the lamen) of the car rotating around the corner, thus the more mechanical trail you have the harder is will be to steer around a corner and it won't just be trying to self centre it will be trying to follow the load path thus steering in the opposite direction! It will also be a nightmare to resist side loading from side wind and bumps. With with KPI on the other hand the it self centres to the chassis via gravity and the centripetal acceleration acts directly on the centre of hub so it is not trying to steer the car, the increased resistance you feel is from the increased load on the outside wheel working harder to centre the wheel because of KPI.
To try and sum up in a sentence Mechanical trail self centres to the direction of the load acting on it (cornering force, wind, bumps etc) but KPI self centres to the cars chassis so it will always try to go straight, whereas a shopping trolley will try to go where other outside factors are pushing them
When I was a wide eyed design student we read the books and weighed up the pro's and cons and decided that we knew better and had new ideas.... and for every idea there was always an example of why it didn't work. For instance placing the IRC under the ground seems like a great way to counteract the CoG and stop the car leaning in a corner, in theory it works, in reality it has been tried to death and it always causes odd handling characteristics.
As for lack of progress? I'm sure Adrian Newey or Ross Brawn would love to be told how they have made no progress.... oh wait.
Thanks thats a very good answer with some good technical input thank you.
I would like to respond by saying that in my oppiniun in a very lightweight road car which will not be experiancing massive lateral acceleration that i am willing so sactifice some steering weight for the beinfits of the camber change due to caster rarther than KPI. my philosophy has always been to keep the tyre vertical for the longest duration possible.
In a 600Kg car with a 40F 60R weight distribution exposed to a 1G lateral acceleration with a small amout of mechanical trail (aprox 30mm) the steering effort should be perfectly managable
tilly
quote:
Originally posted by Doctor Derek Doctors
To try and sum up in a sentence Mechanical trail self centres to the direction of the load acting on it (cornering force, wind, bumps etc) but KPI self centres to the cars chassis so it will always try to go straight, whereas a shopping trolley will try to go where other outside factors are pushing them
As for lack of progress? I'm sure Adrian Newey or Ross Brawn would love to be told how they have made no progress.... oh wait.
quote:
Originally posted by Doctor Derek Doctors
As for lack of progress? I'm sure Adrian Newey or Ross Brawn would love to be told how they have made no progress.... oh wait.
quote:
Originally posted by v8kid
quote:
Originally posted by Doctor Derek Doctors
As for lack of progress? I'm sure Adrian Newey or Ross Brawn would love to be told how they have made no progress.... oh wait.
I've lost the plot a bit here that was the very point I was trying to make - unless we try something different there will be no progress - anyhow no matter
Cheers!
quote:
Originally posted by v8kid
quote:
Originally posted by Doctor Derek Doctors
To try and sum up in a sentence Mechanical trail self centres to the direction of the load acting on it (cornering force, wind, bumps etc) but KPI self centres to the cars chassis so it will always try to go straight, whereas a shopping trolley will try to go where other outside factors are pushing them
As for lack of progress? I'm sure Adrian Newey or Ross Brawn would love to be told how they have made no progress.... oh wait.
Nice summary thanks. Easy to see the argument about KPI when you put it that way and I can see the point now.
Can you do the same for "there are very few cars were the wheel hub centre is not on the centreline (between the pivots) of the upright" please? Certainly thats true but are there could be considerations other than handling ones that cause it. Why is it a no no after all it seems to comply with your points already made.
quote:
Originally posted by Doctor Derek Doctors
As for lack of progress? I'm sure Adrian Newey or Ross Brawn would love to be told how they have made no progress.... oh wait.
I've lost the plot a bit here that was the very point I was trying to make - unless we try something different there will be no progress - anyhow no matter
Cheers!
Ta - Good series of posts here it has clarified my thoughts quite a bit.
Cheers!
With a light front end you need a fair ammount of caster and kpi if you dial in caster alone you end up with a silly ammount of negative camber at
full lock.
If you look at the Lotus Elise forums you will find that although the Elise comes with a fair ammount of both caster and kpi as standard owners are
adjusting the suspension to give more kpi and caster to get sharper turn in.
As to Hillman Imp front ends the swivel hub casting was arranged so the king pin passed about 25mmmm ahead of the stub axle centre, this had the
advantage of allowing room for the cable drive for the speedo to pass through the stub axle.
Imps used a lot of both caster and KPI, istr the caster was about 11 degrees and on a racer the kpi about 7-9 degrees depending on camber, this
combination would result in too much negative camber on hairpins so quite often the leading wishbone pivot point was lowered 25mm giving anti-dive
and reducing caster.
Imp handling when lowered was nothing short of fantastic BUT the driving experience in motorway cross winds more like tacking in a sailing boat
than driving most owners always carried something heavy in thre friont boot to ballast the front and move the cg forward of the centre of
pressure.
[Edited on 14/4/11 by britishtrident]
OK
thank you all for your input it has been invaluable as i am no expert.
Just to clear a few points up:
I can see the value of some KPI to prevent the effects on the steering from crosswinds and road defects.
A large amount of negative camber on the outside wheel under full lock dosnt bother me at all since under normal driving conditions this scinario will
never happen, only when doing low speed opperation such as parking will it become relevelnt.
I am not trying to achive centerpoint steering and would like a small amount of scrub radius.
I have indeed read the book a link was provided to along with many others ( the problem i find is that information in one book contradicts another and
finding out which is correct/best can be difficult)
I am coming round to your way of thinking that a small amount of KPI should be included to prevent the weather cocking and aid self centering. the
explanation by DDD was very good. I am still however reluctant to include a large KPI angle due to its adverse camber effects on the outside wheel, I
am thinking after some very quick maths that an angle of about 3deg would not sacrifice to much outer wheel control, how does this sound?? this would
also bring my scrub radius down to the 10mm range
thanks again
tilly
Sounds like compromise is the order of the day! Would that take your caster up from 7 to around 9 then? Back to the original question then trail will
be around 45 -50mm without stub axle lead.
[Edited on 14-4-11 by v8kid]
I hate compromise i like to acchive perfetion with evrything but this is one of those things that it just dosnt seem to to posible there has to be a
compromise like you say, i had actualy considered bringing the caster down to 5.5 with 0 kpi but with 3 kpi i think 6-7 caster might work il have to
have a play with it late when i get home
tilly
compromise GRRRRR why carnt it be perfect lol
Lotus Elise ( some variation on specs found on the web)
Front suspension
roll centre height 30mm
travel 50mm bump / 60 mm rebound
camber gain in bump 0.31 degrees per inch
frequency 90cpm
*******KPI 12.0 degrees**********
Castor 4.25 degrees
Trail 4mm
Scrub radius 10.5mm
from most examples i've seen, when you add the castor and kpi angles together you get 15 (or very close) so if you go for 3 degrees kpi, expect about 12 degrees castor. thats not an exact rule, but will be about that
quote:
Originally posted by tilly819
I am still however reluctant to include a large KPI angle due to its adverse camber effects on the outside wheel
quote:
Originally posted by Doctor Derek Doctors
quote:
Originally posted by tilly819
I am still however reluctant to include a large KPI angle due to its adverse camber effects on the outside wheel
I am struggling to understand this statement.
You will need a camber increase in cornering, this goes without saying. If you are designing the car then it is a simple job to design the inner wishbone mounts in a position to give the correct camber change for any amount of KPI. You could even have 8° of KPI and have no camber change at all.
Off the top of my head (I'm not the suspension leader and dont have the CAD at my fingertips) our RGB Race car will have about 7-9° of KPI, my personal race car has Cortina uprights with only 4.25° of KPI which is the 1st thing that will be changed to give more.
No DDD to do this you end up with a swing arm length so small it makes the Imp look positively modern! Also the scrub introduced by the track change
becomes very significant.
Compromises again but that's too far in the opposite direction for me!
Cheers!
Oops crossed posts again
[Edited on 14-4-11 by v8kid]
Short virtual swing axle is no problem provided the roll centre is low enough and the roll stiffness is mainly generated from the oposite end of the vehicle.
quote:
Originally posted by v8kid
No DDD to do this you end up with a swing arm length so small it makes the Imp look positively modern! Also the scrub introduced by the track change becomes very significant.
Compromises again but that's too far in the opposite direction for me!
Cheers!
Oops crossed posts again
[Edited on 14-4-11 by v8kid]
quote:
Originally posted by tilly819
quote:
Originally posted by Doctor Derek Doctors
quote:
Originally posted by tilly819
I am still however reluctant to include a large KPI angle due to its adverse camber effects on the outside wheel
I am struggling to understand this statement.
You will need a camber increase in cornering, this goes without saying. If you are designing the car then it is a simple job to design the inner wishbone mounts in a position to give the correct camber change for any amount of KPI. You could even have 8° of KPI and have no camber change at all.
Off the top of my head (I'm not the suspension leader and dont have the CAD at my fingertips) our RGB Race car will have about 7-9° of KPI, my personal race car has Cortina uprights with only 4.25° of KPI which is the 1st thing that will be changed to give more.
When the wheel is steered KPI pushes the outside wheel in a positive camber direction *BAD*
I agree that this can be countered by inter pivot point position to create a negative camber movement however this makes the camber curve more "curved" under bump conditions or more relevantly in the case of the front suspension the breaking condition the wheel will be loosing camber (going negative) due to suspension compression thus making the tyre no longer flat to the road and reducing available breaking grip.
cornering grip is a trade off against acceleration/breaking grip or at least thats the conclusion i have come to
also you will more than lightly find that if you design a system that has a KPI of 8 deg as you say and no camber change due to the inboard points the lateral scrub would be very large
[Edited on 14/4/11 by tilly819]
Having had a nights sleep and re-read the entire thread threw this morning i realise that i may have come accross a little strong in my last post to
DDD so i am sorry of this cause any offence it was not ment to. All of your input has been greatfully recived and i can now see the value of the king
pin inclination.
Put in simple terms I was wrong simple as that.
Now i after doing some more thinking (dangerous i now) i can see no problem designing a system for KPI and caster. this still however leaves the
question of trail. In an eirler post by british trident the lotus elise geo was listed now this has a trail of 10mm therefore with a caster 4.25deg
and a tyre radius of APROX 290mm (about 195/50/15 or 205/50/15) this must use a king pin lead of about 10mm to accive that trail of 10mm or ease the
trail would be about 21mm this seems like a good idea, what are people thoughts????
thanks tilly
[Edited on 15/4/11 by tilly819]
quote:
Originally posted by tilly819
Having had a nights sleep and re-read the entire thread threw this morning i realise that i may have come accross a little strong in my last post to DDD so i am sorry of this cause any offence it was not ment to. All of your input has been greatfully recived and i can now see the value of the king pin inclination.
Put in simple terms I was wrong simple as that.
Now i after doing some more thinking (dangerous i now) i can see no problem designing a system for KPI and caster. this still however leaves the question of trail. In an eirler post by british trident the lotus elise geo was listed now this has a trail of 10mm therefore with a caster 4.25deg and a tyre radius of APROX 290mm (about 195/50/15 or 205/50/15) this must use a king pin lead of about 10mm to accive that trail of 10mm or ease the trail would be about 21mm this seems like a good idea, what are people thoughts????
thanks tilly
[Edited on 15/4/11 by tilly819]
Oh and as for the mechanical trail, I would just let it be a product of the caster angle and tyre size, it seems to be one of those happy coincidences
that doing it that way seems to work just fine for virtually all types of car.
One thing you also need to take into account is 'rake' you may end up with X° of Caster but with 2° of 'rake' in the car the
caster will in reality be X-2° and so your true mechanical trail will be reduced.
If you figure 2° of rake into your system it reduces the trail to 10.8mm anyway, this is why most publications recomending a caster angle higher than
you need to compensate for the rake of the chassis.
Personally I would start with at least 7° of caster (reduced to 5-6° with rake) giving a trail of.... 25.4mm.... ooooh that a nice number for an OCD
sufferer like myself.
quote:
Originally posted by v8kid
Check out this link to a forum with quite a respected contributor on the subject
http://www.eng-tips.com/viewthread.cfm?qid=113181&page=8
Cheers!
quote:
Originally posted by ffrgtm
quote:
Originally posted by v8kid
Check out this link to a forum with quite a respected contributor on the subject
http://www.eng-tips.com/viewthread.cfm?qid=113181&page=8
Cheers!
Please note that in your link he is talking about a strut type suspension. In a strut type suspension, the higher SAI (steering axis inclination... believe this is what you are calling kpi) the more the wheels camber OUT as you turn the steering wheel. Basically the opposite effect of caster (in this regard).
I came across this issue when noticing all of these people in their mcpherson strut cars (stis, evos, m3s) would throw adjustable camber plates on the top of the struts. They would crank them way in and then not understand why their cars actually had worse camber in corners (looking at photos people had taken ect.). I came to the conclusion that you can not use camber plates to adjust strut type cars and must use the eccentric bolts unless you want to lose your dynamic (in the sense of steering angle) camber.
I hope you aren't trying to put a mcstrut type suspension on a locost though.... should really use sla (double a-arm)
Hi chaps,
McP is really the devil in disguise
However if you read the discussion in the link it is generic and not specific to struts
Have to go only 4 days to my dissertation cut off date!
Cheers
quote:
Originally posted by v8kid
Hi chaps,
McP is really the devil in disguise
Cheers