
Tire deflection and frequency
#1
Posted 21 July 2009 - 13:01
Advertisement
#2
Posted 21 July 2009 - 23:12
How much attention do we pay to tire deflection with regard to semi-serious weekend racing? I've been giving this subject a little more thought lately. It seems that grip and wear is a priority but, and yes I'm giving myself a headache again, if the tire's frequency is not sympathetic to the car as a whole wouldn't we expect grip and wear to fall off more rapidly?
I may be wrong, but my first thought is that the natural frequency of the tyre (contact patch in bump/rebound) would be very high compared to other compliances in the car, so tuning is not really the issue. Rather, the tyres should be a compromise of the highest possible natural frequency combined with the highest possible compliance.
#3
Posted 22 July 2009 - 02:32
For tires of roughly the same size and purpose and pressure I don't think they change much.
#4
Posted 22 July 2009 - 16:32
For the limited amount of data I've seen the resonant modes on most racing tyres are above the wheel-hop frequency.
Ben
#5
Posted 23 July 2009 - 00:08
One area I've come across is running the dampers so stiff that the sprung mass is actually underdamped because it's basically deflecting only in the tyre.
That's called boulevard ride and is why soft shock absorber bushes can result in more shake than hard ones, and why we get uppity about friction and hence side loads in shocks. It is also why you can get good shake performance from low profile tires - the stiffer sidewall helps crack the shocks, so they start working earlier. However tires designed for incompetent suspensions throw this away by using an especially compliant sidewall on low profile tires, thereby giving a tire that is functionally much the same as its high profile cousin, but it looks low profile.
#6
Posted 23 July 2009 - 12:57
Greg, Ben,
Would you expect a low profile tire conctructed for an "incompetent suspension" to feel lazy when matched to a competent suspesnion? I realize this is a more or less boundless theoretical question - no exact numbers to play with.
Ben, I think I am observing the same thing, though the esoterics are beyond me.
#7
Posted 23 July 2009 - 15:25
I beleive several cars have had "special" tires including IIRC the original Elise. If the OEM tire was special I assume the "not quite the same"! aftermarket version will compromise the suspension/damper settings.
Also I have seen claims that wheels with hollow spokes ( i.e the spoke interior is directly connected to the tire chamber "give a softer ride" with low profiles. Not sure if this is actually true but if so it would suggest the natural tire frequency is significant for ride. As the volume of the spokes must be smal versus the normal tire internal volume I don't see why there is an impact with hollow spokes but it has been claimed
#8
Posted 24 July 2009 - 01:11
"incompetent suspensions and tires" make me think again about tire makers supplying special tires for specific OEM applications and then never making the exact same tire available in the aftermarket supply chain.
I beleive several cars have had "special" tires including IIRC the original Elise. If the OEM tire was special I assume the "not quite the same"! aftermarket version will compromise the suspension/damper settings.
Also I have seen claims that wheels with hollow spokes ( i.e the spoke interior is directly connected to the tire chamber "give a softer ride" with low profiles. Not sure if this is actually true but if so it would suggest the natural tire frequency is significant for ride. As the volume of the spokes must be smal versus the normal tire internal volume I don't see why there is an impact with hollow spokes but it has been claimed
Your first point is wrong, the OEM specs are usually set so that a %age of tires are rejected (typically 15% or more) . These could be scrapped, but in practice of course they are sold as aftermarket tires. They are the same as the OEM fit tire, they just might not meet some of the specs.
Your second point is right, in fact almost any vehicle that is sold >10000 pa will actually have its own tire. I agree, replacing the original fitment tire with a different make or model is a gamble. AFAIK all Lotusses for a long time had their own tires, this caused problems with the fat bottomed girl Elan as it was developed for a non GY tire, and GY struggled to duplicate its performance, we had to use GY because of F1. I think in the end the suspension was retuned around the GY tire.
Yes wheel stiffness is important subjectively, but it is very hard to see where the additional compliance really affects the system.
#9
Posted 24 July 2009 - 01:31
I think the reference is to increased air volume (in the hollow spokes) leading to increased pneumatic compliance.Yes wheel stiffness is important subjectively, but it is very hard to see where the additional compliance really affects the system.
Edited by gruntguru, 24 July 2009 - 01:32.
#10
Posted 24 July 2009 - 02:40
#11
Posted 24 July 2009 - 04:19
#12
Posted 24 July 2009 - 13:12
I think the reference is to increased air volume (in the hollow spokes) leading to increased pneumatic compliance.
#13
Posted 29 July 2009 - 11:54
After reading above I can see how there'd be some small changes in the suspension frequency with different tyre pressures - What I'd really like to see is a few damper histograms on the same car but with different tyre pressures. The problem with that is getting a tyre that still works over a good enough range of pressures and I suspect you'd be hard pressed to see any chages ..... but I'd like to know for sure.
#14
Posted 29 July 2009 - 14:33
Edited by meb58, 29 July 2009 - 14:35.
#15
Posted 29 July 2009 - 14:59
#16
Posted 29 July 2009 - 18:41
#17
Posted 29 July 2009 - 19:26
What'st the point of air pressure adjustment to affect car handling? Is it a spring rate thing or a weight balance thing?
It's mainly a spring rate thing. There's a lot of discussion about "optimum" tyre pressures but the performance is similar over a relatively wide range if it's a stiff tyre - i.e. GT. I've not got as much experience with softer carcass tyres such as single seaters and by definition a less stiff case will be more pressure sensitive.
In virtually all the situations I've been involved with tyres going high in pressure due to incorrect cold pressures or a sudden rise in ambient temp is rarely a problem if they rise equally front to rear. If the rears rise higher than the fronts (common in powerful RWD cars) the car will definitely go O/S and the driver will be upset. Moral of this story being, if you take a little bit of grip they won't really notice, if you change the balance even slightly you'll hear about it pretty fast.
Ben
#18
Posted 30 July 2009 - 03:26
I vaguely remember that if you increase the pressure by 20% you gain 10% in vertical rate, for a typical small production tire.
Base rate is say 220 N/mm at 30 psi
Quite how you turn that into a shocak absorber curve I leave for you!
#19
Posted 30 July 2009 - 10:11
Well somewhere in my sekret files I have a plot of tire vertical rate vs pressure.
I vaguely remember that if you increase the pressure by 20% you gain 10% in vertical rate, for a typical small production tire.
Base rate is say 220 N/mm at 30 psi
Quite how you turn that into a shocak absorber curve I leave for you!
For the race tyres I'm working on at the moment the percentage increase in rate is similar to the percentage change in air pressure.
Ben
Advertisement
#20
Posted 10 August 2009 - 13:02
Edited by meb58, 10 August 2009 - 13:14.
#21
Posted 11 August 2009 - 11:02
From a production car perspective, do engineers study the tire, damper, spring as a unit/system with sprung mass? I assume that some asumptions can be made based upon experience...and that the upper end of the sporting spectrum might warrant some expensive analysis...?
Road vehicles are much more complex than race vehicles and, for what it is worth, I have concluded that suspension set-up for such vehicles is usually compromised by NVH issues. In no particular order, vehicle properties that affect set-up directly or indirectly include:
- vertical tyre stiffness
- the ability of a tyre to "wrap around" road obstacles (not sure what to call that - circumferential stiffness, perhaps?)
- damper isolators (top mounts)
- power train suspension (&, I suppose, sub-frame suspension when fitted)
- seat compliance and damping
- a range of damper characteristics
- long bump rubbers (often used at the front axles for, perhaps, misguided reasons)
- indirect "noise" paths.
Single axle models are simple to implement and are used widely, but they can be misleading.
Undoubtedly (in my biased view) the best way of understanding set-up issues is by whole vehicle rig tests. These cannot be used to set-up a vehicle completely, but can be used to put settings in the correct ball park, & can usually save much time by revealing the source of perceived problems.
Edited by DaveW, 11 August 2009 - 11:06.
#22
Posted 11 August 2009 - 11:50
- the ability of a tyre to "wrap around" road obstacles (not sure what to call that - circumferential stiffness, perhaps?)
Cuddle factor?
#23
Posted 11 August 2009 - 12:08
Road vehicles are much more complex than race vehicles and, for what it is worth, I have concluded that suspension set-up for such vehicles is usually compromised by NVH issues. In no particular order, vehicle properties that affect set-up directly or indirectly include:
- vertical tyre stiffness
- the ability of a tyre to "wrap around" road obstacles (not sure what to call that - circumferential stiffness, perhaps?)
- damper isolators (top mounts)
- power train suspension (&, I suppose, sub-frame suspension when fitted)
- seat compliance and damping
- a range of damper characteristics
- long bump rubbers (often used at the front axles for, perhaps, misguided reasons)
- indirect "noise" paths.
Single axle models are simple to implement and are used widely, but they can be misleading.
Undoubtedly (in my biased view) the best way of understanding set-up issues is by whole vehicle rig tests. These cannot be used to set-up a vehicle completely, but can be used to put settings in the correct ball park, & can usually save much time by revealing the source of perceived problems.
Good list Dave.
One thing that has often occurred to me is why race cars don't look at preloaded top mounts instead of solidly locating the dampers with spherical bearings. There are lots of "force blow-off" dampers - surely a preloaded top mount for the whole coilover unit would be as effective and potentially more field adjustable?
Ben
#24
Posted 11 August 2009 - 12:47
Are you referring to a Pillow Ball as the solid mount? I must admit, although I know what this is, and have used a few and can tallk about them, I do not know where the term Pillow Ball came from...I have an idea or two but niether is fit for print. By pre-loaded do you mean some sort of spring steel as the actual perch? This observation is in the Mac Strut direction.
Greg,
Regarding a saloon type track car...I guess testing and more testing is really the best way to assess what works and what doesn't. I have to form a base line somewhere and use that as a point of departure - either better or worse as tire type or damper characterisitcs change etc. Makes the process more interesting to my mind in any event.
#25
Posted 11 August 2009 - 13:15
Ben,
By pre-loaded do you mean some sort of spring steel as the actual perch? This observation is in the Mac Strut direction.
What I'm envisaging is a spring (or bump rubber) compressed by a certain amount. If the spring is 100N/mm rate and you compress it by 5 mm it will take 500N to make it deflect any more. The idea being you set the preload above the highest steady state force you a likely to achieve so that the mount doesn't deflect. If you have a shock load (e.g. kerb strike) that exceeds the preload the mount then deflects and your total rate becomes the 100N/mm in series with the main spring - i.e. a lower rate and you reduce transmissibility over kerbs without making the thing too soft overall.
The damper manufacturers have been doing this since the Mid 90s - i.e. including a force based blow-offs to improve kerb strike behaviour, but Mr Williams will probably know more about this than me.
Ben
#26
Posted 11 August 2009 - 15:24
If I missed your idea again I'll sit back for a while and read only...
#27
Posted 11 August 2009 - 18:30
One thing that has often occurred to me is why race cars don't look at preloaded top mounts instead of solidly locating the dampers with spherical bearings. There are lots of "force blow-off" dampers - surely a preloaded top mount for the whole coilover unit would be as effective and potentially more field adjustable?
Ben
Mmm..
Non-preloaded top mounts render dampers less effective than they might be, & can reduce hub mode damping ratios dramatically (because the hub can move with no damper movement). Generally, one requirement for good mechanical control is to minimize compliance acting in series with the dampers. Damper blow-off is normally used to limit the maximum damper fluid pressure to a preset value. It is used to limit how far a sprung mass will be launched by a large discrete input & has proved to be a valuable control for both aero (particularly) and non-aero vehicles on rough race tracks.
I have never experienced preloaded top mounts on a race vehicle. but I could imagine that they might prove beneficial under some circumstances. One note of caution is that top mounts will store energy when they move, most of which will be returned & must be dissipated by the dampers at a later time. That is not be the case for a blow-off damper.
#28
Posted 12 August 2009 - 00:25
Mmm..
Non-preloaded top mounts render dampers less effective than they might be, & can reduce hub mode damping ratios dramatically (because the hub can move with no damper movement). Generally, one requirement for good mechanical control is to minimize compliance acting in series with the dampers. Damper blow-off is normally used to limit the maximum damper fluid pressure to a preset value. It is used to limit how far a sprung mass will be launched by a large discrete input & has proved to be a valuable control for both aero (particularly) and non-aero vehicles on rough race tracks.
I have never experienced preloaded top mounts on a race vehicle. but I could imagine that they might prove beneficial under some circumstances. One note of caution is that top mounts will store energy when they move, most of which will be returned & must be dissipated by the dampers at a later time. That is not be the case for a blow-off damper.
Tuning top mounts is a pain, there is remarkably little rhyme or reason to it. The reason for that is probably that everything affects everything - a stifffer bush will transmit more high frequency and low amplitude stuff, and help crack the friction, which is good, but is likely to have a stiffer coning rate, so it puts more sideload into the shock, so friction increases, which is bad.
The other problem is that ride and secondary ride are incredibly hard to measure in any meaningful way that correlates with subjective impressions. I was looking at atest report where they increased tire pressure by 50%, the change in the ride measurement they were using was around 2%.
#29
Posted 12 August 2009 - 07:44
Tuning top mounts is a pain, there is remarkably little rhyme or reason to it. The reason for that is probably that everything affects everything - a stifffer bush will transmit more high frequency and low amplitude stuff, and help crack the friction, which is good, but is likely to have a stiffer coning rate, so it puts more sideload into the shock, so friction increases, which is bad.
The other problem is that ride and secondary ride are incredibly hard to measure in any meaningful way that correlates with subjective impressions. I was looking at atest report where they increased tire pressure by 50%, the change in the ride measurement they were using was around 2%.
I believe we are talking about different animals, Greg. Ben & I were discussing the possibilities of using top mounts in race vehicles.
In the broader context of road vehicles, top mounts are extraordinary in many ways, but they are sometimes responsible for uneasy set-up compromises, largely because their effect is not understood or - more accurately - because the shorthand description of their function can be misleading. Your statement "a stifffer bush will transmit more high frequency and low amplitude stuff" is an example. I know what you mean (..will transmit more of a frequency), but the statement you made could lead to the conclusion that softer top mounts will cure all ills.
I recall being invited to rig test one troublesome development vehicle. My conclusion that the top mount stiffnesses should be doubled was greeted with some derision, to be followed a week later by the admission that increasing (actually tripling) the stiffness of the top mounts had transformed both handling and subjective ride. The rig test provided (in this case) both the rhyme and the reason for modifying the top mounts. If I have your allegiance correct, I think you already have a four post rig available to you that could easily be transformed into a useful cost saving quantitative tool (If I'm correct, then I did it, on a temporary basis, some years ago to help set up a "Super V8").
#30
Posted 12 August 2009 - 09:55
If I have your allegiance correct, I think you already have a four post rig available to you that could easily be transformed into a useful cost saving quantitative tool (If I'm correct, then I did it, on a temporary basis, some years ago to help set up a "Super V8").
It's no secret, I work on Ford of Australia Falcon platform, mostly, running ADAMS and trying to impose/utilise some analytical approach on the chaotic Vehicle Dynamics development process. But I do slum it assessing real cars as well. The 4 poster is very busy, on other, bigger, programs.
#31
Posted 12 August 2009 - 12:35
#32
Posted 13 August 2009 - 09:56
It's no secret, I work on Ford of Australia Falcon platform, mostly, running ADAMS and trying to impose/utilise some analytical approach on the chaotic Vehicle Dynamics development process. But I do slum it assessing real cars as well. The 4 poster is very busy, on other, bigger, programs.
Nothing wrong with Adams (apart from size & efficiency). In my (very biased) view, simulations are useful but, ultimately, only repackage what you know &/or assume about a vehicle. Rig tests are more limited in many ways, but they do have the advantage that they can yield information about a vehicle that was previously unknown (possibly why I have a steady trickle of F1 customers). The ideal would be to have an Adams model that has been validated (adapted) using information & properties extracted from rig tests. I would suggest that you could build a technical & financial case for a share of rig time. It is possible that the other programs might benefit from that in the longer run.
#33
Posted 13 August 2009 - 11:27
Nothing wrong with Adams (apart from size & efficiency). In my (very biased) view, simulations are useful but, ultimately, only repackage what you know &/or assume about a vehicle. Rig tests are more limited in many ways, but they do have the advantage that they can yield information about a vehicle that was previously unknown (possibly why I have a steady trickle of F1 customers). The ideal would be to have an Adams model that has been validated (adapted) using information & properties extracted from rig tests. I would suggest that you could build a technical & financial case for a share of rig time. It is possible that the other programs might benefit from that in the longer run.
Oh I'm no great fan of the complexity of ADAMS, most times for a passive car a bicycle model , or 4 wheel bicycle would tell you the same thing, if you can decide what the appropriate characteristics are. But for an organisation where changes are made on a component basis it substitutes cheap CPU time for expensive IQ time. Once you add EPAS into the equation then the bicycle model becomes less helpful. Bear in mind that 50% of an ADAMS runtime is tire code, 25% is steering, so all the perceived 'complexity' is just the last 25% of the whole model.
#34
Posted 13 August 2009 - 13:09
I know nothing about modeling such things but I imagine there might be some very interesting data between the two examples above.
#35
Posted 13 August 2009 - 18:15
So from the perspective of being effective, and not necessarily efficient, do we analize the difference between a rod end upper perch and a more malleable perch material as the strut precesses the perch?
Apologies for ignoring your posts meb58, but your questions are difficult to answer in words. Here is an attempt:
The effect of a "top mount" (polymer compliance placed in series with a damper) can, perhaps, be understood by thinking of a damper as a "spring" whose "stiffness" increases with increasing frequency. When a top mount has minimal compliance, the damper will be forced to move when its hub moves relative to the sprung mass, & so it will dissipate energy consistently up to a high frequency. However, when the top mount is compliant, the attached damper will be allowed to "lock" at high frequencies, so the damper will dissipate ever less energy per unit hub movement (relative to the sprung mass) as the frequency increases. Ultimately, if a top mount is sufficiently compliant (or the damper sufficiently "strong"), control will be lost over hub oscillations & hence over contact patch load variations. Both handling and "NVH" will be affected adversely by top mounts that are too compliant.
The above means that polymer top mounts are not normally used for race vehicles where "NVH" is, at worst, an irritation, whilst control is a primary objective.
On the other hand, spherical bearings are not used as top mounts for road vehicles because any high frequency "noise" generated or transmitted by a damper tends to be passed directly to the sprung mass. Any bearing free play will also generate "noise". Both will have a damaging effect on "NVH".
As always, a compromise is required for top mount stiffness for road vehicles, and a good target is to choose a top mount with a working "rate" no less than tyre stiffness times 3. Some might say that ratio is conservative, but it worked well when designing active suspension installations.
I'm not sure what a "pillow ball" is, but it would probably not be acceptable unless it was capable of carrying tensile loads (e.g. with a wheel off the ground).
I hope you find this helpful.
#36
Posted 13 August 2009 - 23:46
a good target is to choose a top mount with a working "rate" no less than tyre stiffness times 3. Some might say that ratio is conservative, but it worked well when designing active suspension installations.
That is an interesting way of thinking about it, remember to take motion ratio into account. That's a bit low in my opinion, for normal shocks, on an X5 sort of vehicle with SLAs/multilinks.
#37
Posted 14 August 2009 - 06:24
That is an interesting way of thinking about it, remember to take motion ratio into account. That's a bit low in my opinion, for normal shocks, on an X5 sort of vehicle with SLAs/multilinks.
You're right. I should have mentioned motion ratio - I think of everything happening at the wheels, anyway.
The logic behind the stiffness ratio is based on hub mode natural frequencies damper off compared with damper locked. The maximum possible hub damping ratio reduces as the interval between those two asymptotes reduces, and the reduction in damping ratio becomes greater as damper settings are increased or decreased from the "maximum" settings. Neglecting the parallel spring, x3 gives a factor of 2 between the two natural frequencies, which provides some scope for achieving both reasonable hub damping and some control over the sprung mass modes.
I am happy to hear you say my "rule" is low. You would be surprised (perhaps) by the number of production road vehicles that suffer from top mounts that are too soft. One reputable marque has a (rear) top mount with a lower rate than the corresponding tyre. Its vehicles have no rear hub mode control worth speaking about. I imagine that is because the set-up "style" was set to solve a specific problem when tyre rates were much lower, & they have relied on a subjective set-up culture since. Again, "boy racer" versions of existing production vehicles usually involve fitting higher performance (& usually higher stiffness) tyres. But I have yet to see one with revised top mounts....
#38
Posted 14 August 2009 - 14:41
I never thought about tensil loads...and yes, thank you, very helpful. Your description also magnifies, as Greg wrote, why friction in the damper and perhaps other parts is so important.
I apologize for writing questions that do not follow engineering diction...this is not my field and I feel fortunate to be able to ask you folks questions. I appreciate the feedback!
Apologies for ignoring your posts meb58, but your questions are difficult to answer in words. Here is an attempt:
The effect of a "top mount" (polymer compliance placed in series with a damper) can, perhaps, be understood by thinking of a damper as a "spring" whose "stiffness" increases with increasing frequency. When a top mount has minimal compliance, the damper will be forced to move when its hub moves relative to the sprung mass, & so it will dissipate energy consistently up to a high frequency. However, when the top mount is compliant, the attached damper will be allowed to "lock" at high frequencies, so the damper will dissipate ever less energy per unit hub movement (relative to the sprung mass) as the frequency increases. Ultimately, if a top mount is sufficiently compliant (or the damper sufficiently "strong"), control will be lost over hub oscillations & hence over contact patch load variations. Both handling and "NVH" will be affected adversely by top mounts that are too compliant.
The above means that polymer top mounts are not normally used for race vehicles where "NVH" is, at worst, an irritation, whilst control is a primary objective.
On the other hand, spherical bearings are not used as top mounts for road vehicles because any high frequency "noise" generated or transmitted by a damper tends to be passed directly to the sprung mass. Any bearing free play will also generate "noise". Both will have a damaging effect on "NVH".
As always, a compromise is required for top mount stiffness for road vehicles, and a good target is to choose a top mount with a working "rate" no less than tyre stiffness times 3. Some might say that ratio is conservative, but it worked well when designing active suspension installations.
I'm not sure what a "pillow ball" is, but it would probably not be acceptable unless it was capable of carrying tensile loads (e.g. with a wheel off the ground).
I hope you find this helpful.
Edited by meb58, 14 August 2009 - 17:58.
#39
Posted 14 August 2009 - 18:01
A pillow ball - so it's called -
I'v always called them spherical strut tops and Google agreed with me....
http://www.eurospeed...au/strutTop.htm
http://www.google.co...s...mp;oq=&aqi=
Advertisement
#40
Posted 14 August 2009 - 19:27
I'll go with sperical top strut. Thx!
#41
Posted 15 August 2009 - 11:01
I am guessing that it is called pillow as the support it goes on may be called a bed.
#42
Posted 15 August 2009 - 11:33
And to be clear, I know a lot less than you folks.
Impossible, how can you know less than nothing?

#43
Posted 17 August 2009 - 12:55

#44
Posted 28 December 2011 - 15:15
At the end of this painful video (6:50) there is a great shot of a GP2 car oversteering mid to exit on a corner and showing a stange 'vibration'. At first I thought it was a tyre issue with the Left Reat tyre, but then when the car 'flicks' and the LR tyre become the loaded side the RR tyre develops the same response.
#45
Posted 28 December 2011 - 16:24