IRS WHEEL TRAMP, CONTINUED 2

[Ron364]

thread800-286119
As an update to then original thread on this IRSweheel tramp [Americans incorrectly call IRS tramp as hop}. Live axles do indeed hop vertically, but irs tramp does not actually lift the driven wheels, hence tramp is more accurate.
Update on my experiments and I really hope Greg and Pat and other previous contributors read this and offer comment.
Greg, as a followup to engine torque and the mounts being a source, well after doing extensive gopro filming of all of the likely energy storage bits; engine, gearbox, diff and wheel joints, i found that as expected they all tense up and release energy when wheelspin occurs. No wheelspin = no opportunity to release energy = no tramp. At tramp, the engine rocks the most on its liquid mounts, likewise the gearbox is bolted to the engine and of course rocks to the samne frequency, about 8 Hz. Greg, i fitted a shocker to the engine, mounting it horizontally using one of the few shockers that can work horizontally. I used 2000 QUAD Mustang shockers, as other types get an airlock in their tube that limits use. Result was that the resonance in the engine bay was reduced by about 75%, and reduction at the wheel, where it matters, was also greatly reduced.
The most alarming movement was seen in the diff; mounted with two rubber mounts at the front and a single offset mount at the rear. Gopro movie shows massive vertical resonance with minimal sidways movement. Hence a vertical rubber “snubber” would be a good helper here. A BMW fix in usa seems to be a fabricated rod end as a fourth mount with a solid bush in the original 3 diff mounts, hence solid diff bush with std rubber mounts everywhere else in the rear.
Gopro at the wheels showed the same 8Hz resonance, no vertical lift, only slip/grip cycles with some movement as the rubber mounts compressed causing significant toe change.

My mods, all done separately. Engine damper; excellent mod and the factory should have done it.
Polyurethane diff bush in each of the 3 mounts; NHV increased slightly with poly stiffness 90. Tramp surprisingly little changed. BMW “Guy” confirmed that he also failed with poly but succeded with solid joints. My experiments showed that these claims are dangerous because my experience is that only testing on a wet road is by far the best way to test. When questioned, suppliers of various poly bushes or arms or mounts showed that wet road testing had not been done. This certainly applies to claims from usa suppliers.

The use of big/small diameter axles gave unclear results. I used the Camaro use of differing axles as a guide and built my own axles to a similar torsion ratio. Improvement was minimal and resulted in me following up with GM designers. They also found it didn’t fix the tramp problem and ultimately did what bmw and mercedes did, and that is to use traction control to reduce or even stop all wheelspin. The trick here is that TC is not cancelled when pushed and needs either a 10 second hold down time, or simply cannot be turned off completely. How disappointing that these mega companies cannot engineer a solution and use basically the abs sensors to stop wheelspin rather than fix the cause, I call it waving them white flag!

I have systemically changed rubber bushes for poly or rod ends. The Camaro uses nice rod ends in their toe links, and added another spherical joint to the wheel hub. I can only conclude that GM has failed to find a complete solution and the IT solution used by others really only hides the problem. Electric throttles are part of this stop the wheelspin “cure”. Despite the throttle being 100%, the engine cuts to idle positioning!

Conclusion: i think the “

http://www.hopnot.com”

fellow using a “horizontal damper at the rear wheels is offering the best and certainly simmplest solution, and at minimal cost too. He also found that he needed an engine damper for full 100% wet weather solution. However, the layout of the ve cOMMODORE,/G8 is so far, too difficult to mount a similar damper. It needs innovative brackets to fit whereas hopnot solution is for a Chrysler 300 type car, similar rear IRS but different enough to prohibit it on the Commodore/Pontiac G8

So there it is: use an engine damper, fabricate a diff snubber, mount dampers horizontally at each driven rear wheel if possible! Using poly bushes is dangerous as they can only be used when the geometry allows the particular arm to rotate precisely {rotate over each connecting bolt}. This is not the case when you have semi trailing arm setups, like Camaro, C ommodore and Bmw or similar designs. I found poly suppliers simply do not understand how polyurethane works compared to vulcanised original rubber bushes. The rebound frequency of poly is considerably higher than rubber. Poly also does not like being forced to squish at an angle to the bolt in the joint, and will form an oval shaped hole after a few months in service and can resault in considerable dangerous movement in an arm or link.

After extensive testing, i am sticking with my thought that tramp is a system resonance caused by stored system energy being released through the wheel when wheelspin occurs, that’s energy being again stored and then release resulting in broken slip/grip wheelspin as the car moves along the road. If full wheelspin occurs, this bypasses the tramp mode, but of course forward progress is minimal.

Perhaps I should join the white flag crew and use electronics to stop wheelspin. I look forward to reading your comments. Please foregive typos, hard going on my iphone at night. Cheers

 

[LionelHutz]

FYI, the latest 3 generations of Corvette have the differential attached to the end of a 1-piece drivetrain assembly so it essentially creates a long "torque arm". The design also eliminates the issue of engine/transmission torsional movement. They still do it.

If you want to see something interesting, jack up the car, take off the wheels and re-install the nuts to hold the rotors. Then, let the car run in gear and watch the effect on the rear suspension as the traction control applies the rear brakes. You'll likely see the suspension moving quite a bit. Controlling that movement could explain why the hopnot solution helps.

 

[Ron364]

Getting near the end of asking this forum for help on this topic. “Cibachrome” could you please have another go at explaining what you mean re tyre design. The VE Commodore and the Camaro were both designed in Melbourne to use very grippy Bridgestone Pontenza 050 tyres. I don’t like low profile 40 series tyres on road cars, perhaps address this too. The VE rear used a BMW 5 that was dismantled and viewed to the nth degree. It formed the basis of the IRS design despite the knowledge that those cars were generally very poor in ride on typical Aussie roads. Greg, the VE is not an Omega simple lower trailing arm, nor is it a Toyota copy. The top L shaped pseudo wishbone was used to fit available space. The weird pivot planes top and bottom are not parallel to each other and are at about 30 degrees to the centreline. Videos taken by BMR simply show the geometry change arising from the mismatched top and lower “wishbone” planes as the hub moves vertically with some rotation with normal driving, yet the company uses it as a display of massive movement that needs their products to correct! A rear toe correction arm pivots in a different plane to keep toe out changes reasonable. This geometry, especially the use of a toe correction arm, requires some joints to compress at weird angles to allow for the pivot planes not being “along the wishbone bolts”. Using stiff joints like poly or delrin is clearly not appropriate everywhere.

I agree with Greg that the engine mounts are the major source of the problem as the powerful V8 has lots of energy to store in its fluid filled mounts. Similarily, energy is stored in the corresponding gearbox rubber mount. The prop may well twist too, and then to the diff that is rubber mounted to a cradle that is rubber mounted to the monocoque. The onset of tramp occurs when the rear tyres lose traction. The release of the stored energy adds to the engine torque and makes the engine violently resonate, similarly the gearbox as seen by the gear lever resonating inside the cabin. The diff resonates about its only rear mount in a vertical motion, and it is indeed frantic as per my GoPro vision. The cradle moves but really this is minimal in comparison to the diff and much less violent. The wheels do not “hop”, that is a live axle outcome as the irs axles spin in bearings with no rotation being applied to the outer hubs only to the wheels. The tyres grip for say 6 inches, then spin for another 6 inches and this slip/grip process continues until traction is fully regained mainly by the throttle being lifted out of sympathy. The wheels themselves resonate in several horizontal directions seemingly as the arms’ rubber links flex. The massive amount of thumping and banging from tramp must surely have made the designing engineers cringe when testing! Pity they didn’t fix it!

Greg has suggested that “his” Lotus Charlton solution of putting a rubber twin tube in the tail shaft to dampen things actually works. I contacted English owners and they confirmed it too. This solution is beyond a DIY owner, and no manufacturer is interested now that a simple electronic solution is available!

My trial and error experience has shown that my biggest gain was obtained by using a $50 damper in the engine bay. Helped greatly in the dry road tramp, but still present on wet roads. Solid cradle mounts in the rear also helped but did not cure tramp. Poly diff mounts made very little improvement, a fact that the BMW “diff lockout” fellow confirmed to me; he tried poly but he needed solid aluminium diff mounts to achieve any success. I agree with a previous contributing forum member that poly rebounds at a very high rate and can make the situation worse despite reducing movement. Axles; at great expense I made up some big/small high strength axles to a torsion ratio of 4 to 1 like the Camaro and CTSVs; result was only a minor improvement. Despite the advertising hype from GM, I later confirmed that they drew the same conclusion and used electronic TC control; no wheelspin = no tramp. Funny how BMW and Mercedes also reached the same non-engineering solution.

Disappointingly, I have found minimal ideas from this forum despite the purpose of having a forum to share knowledge.

I found that the aftermarket generally has poor knowledge of the cause let alone the solution. When I saw poly bushes being sold with their inner tubes shorter than the poly, I knew they had no idea how their own joints actually worked. Discussion with several firms both in AMERICA and Australia soon made me conclude they were more interested in drag racing traction rather than curing wet road tramp.

The best free thinker on this subject has been “Hopnot”. Treating the end product at the wheels and not the cause at the engine, was a great solution (he does use a simple engine steady though). Cheap $100 dampers at the wheels to dampen the wheel resonance works. Same as putting your finger on a resonating crystal glass. Funny how other manufacturers have spent millions and failed as they never thought of the easy option. Annoyingly, it is too hard, so far, to fabricate mounts on the subject car (Commodore 2007-2017) and the Camaro (2010-2018).

So what has the forum offered? Grippy tyres I think but remain confused by Cibrachrome, rubber prop by Greg, and reference to Mu curves and models by others, being of little actual help. I remain amazed that huge companies like Ford sell Mustangs with IRS that have horrendous tramp. Solutions being left to the aftermarket and as I have found, they don’t test on wet roads and hence I have no faith in their ideas. Hopnot, GM, BMW, Ford and Mercedes would like to speak to you. R

 

[NormPeterson]

http://www.edy.es/dev/docs/pacejka-94-parameters-explained-a-comprehensive-guide/

Norm

 

[cibachrome]

If your question to me is about the usage tire properties related to power hop and tramp, which one of these tires do you think has the likelyhood of a problem ? And the Oz tire selected for the GMX521 program submission were 245/50R19 Yokohama Advant S/T at 32 psi on the Goat but faded to the 'Stones needed for the global market. Too bad your concerns don't show up anywhere on the "Voice of the Customer" mission statement for the G8. The Market clearly doesn't feel your pain I'd say !.

 

[Ron364]

Norm, thanks for the modelling reference. I presume the main manufacturers have access to the same or similar models. The outcome is that their IRS cars all seem to tramp, so does this mean the models are of no help, or perhaps other factors like cost override. I see no value in this model in the situation of working on cars that have already been designed and are in production and tramp. I was hoping to obtain a direction from forum members.

Cibrachrome, thanks for the reply and the reference to the “Goat”, an Australian GMH “Monaro” sold in America as the Pontiac GTO that uses a simple semi trailing arm that Greg incorrectly assumed. The subject car is very different, the only common element being both tramp like crazy. I find it difficult to follow your posts. The subject cars, Australian VE Commodores (the G8 and GXP is the same vehicle except the G8/GXP is sold in America and is left hand drive), and the near identical Camaro since 2010. They use 245/45/18 and 245/40/19 Bridgestone Pontenzas in Australia and a variety of similar grip tyres in America. If you think that I am the only person worried about tramp, then you are mistaken. Ask BMR or gforce1320 for verification. If you have a clear suggestion to solve the problem, perhaps with tyre choice, then please give it in plain English. The Pontenzas are very grippy tyres with stiff sidewalls in the 40/45 profile. A low profile that I personally dislike for road cars. If you know of better road tyres, please list them. The simple rule “no wheelspin = no tramp” holds true, so grip is king. The link below shows the rear end in good detail.

 

[cibachrome]

Not sure what tire resources you have nearby, but I took a look at tire constructions and construction modifications that were going on at the time. Without a doubt, somebody was having trouble with the Bridgestone 245/45R18 and 245/40R19 submissions. (These are all on 8.5" rims).

If you want 'grip', there was a Goodyear and a Continental at the top of the list in the 18" size. In 19", there is a Michelin and a BFG that look interesting. If you want to go off the beaten path, There is a wizzbang Pirelli in 245/50R19 that showed up as the perfect Artgum Eraser. (Lots of real rubber instead of fake rubber.)

Now. Here is a cheap experiment to run on your car that may reveal the answer you seek. What I see in most of these tests is wayward Mx (Tire overturning moment). Not the stiffness so much, but the DC value. It's equivalent to the scrub radius you would think about on a steered upright. Under power, (as in MASSIVE power application) this design 'feature' is a player in torque induced steer (as in break your fingers from steering wheel fight). What I recommend is that you reverse your tires in the rims (outside sidewall now mounted inside). This may reverse the sign of Mx (unless its rotation direction dependent). The tramp may get horribly worse or get much better.Only need to do the rears. Next play, you would add wheel spacers to change the effective rear scrub radius and witness a change in tramp. You won't have access to spacers with negative thickness, but if the phenom is much worse, a change in wheel offset would make sense at this point. Tires and wheels have to fit the wheelhouse, though.

Then report back.

 

[NormPeterson]

Hans Pacejka is one of the top authorities on tire behavior and how to model it. I suggest looking at the mu-slip curve more for its shape and what that might imply. It's more difficult to keep your tires operating within, say, 0.02g of a sharp peak than within the same 0.02g of the top of a wider one with more gradual curvature. I provided the link mostly to point you at how the "C" and "D" terms mentioned by Ciba fit into this.

Here's another Pacejka link that I only had to type his first initial and last name to find.

https://books.google.com/books/about/Tyre_and_Vehicle_Dynamics.html?id=wHlkbBnu9FEC

On "no wheelspin = no tramp".
In a front engine/RWD car with more than half of the static weight being carried by the front tires, engine torque (briefly assisted by sudden clutch engagement) times overall gearing in 1st gear and divided by drive tire rolling radius may be putting you in a situation where wheelspin with reasonably normal road tires simply cannot be avoided. Keep in mind that anti-squat for most IRS systems tends to be significantly lower than what is possible with a live axle system intentionally designed for hard launching. Meaning that you aren't going to get as much traction help from near-instantaneous load transfer to the rear tires. Nor would I expect tires with higher vertical stiffness to cope as well with sudden traction loads as more compliant tires. Wrinkle-wall slicks were developed to serve a very specific purpose . . . and they aren't much good for anything else.

I'm quite aware of the unhappiness over the electronic approach GM has taken in the 6th generation Camaro (Alpha chassis), most noticeably in the supercharged V8 versions fitted with manual transmissions.

Norm

 

[GregLocock]

I was obviously wrong on the suspension geometry. I'd observe that the traction forces are not directly resolved longitudinally but instead go through those red NVH bushes (they provide the longitudinal compliance and damping at the wheel) and the pivots. That setup works well for 150 hp front wheel drives. My experience is that a longitudinal arm (eg Chapman strut) that directly resolves the traction forces is a lot less trouble than messing about with nice packaging friendly designs.

Now as to solutions. I've told you what worked on the Carlton. You complain that an owner can't fit that solution. True-ish. Nonetheless that indicates that at least some manufacturers were aware of the problem and knew how to solve it mechanically. Since these days it can be solved with correctly calibrated traction control then you'll no longer see OEM mechanical solutions for it. That's life. You seem to have found an aftermarket solution. Great, go and fit that.

The reason you are getting little traction here is that there isn't a big bumper book of secret solutions with the answers written down, problems have to be defined and investigated before most solutions can be applied. Different cars may need different solutions. I'd estimate that a proper mechanical resolution of that problem would cost about 1000 hours of engineering time and add about $200 in parts to the cost of the car, or about $1000 on the selling price. That is not small potatoes.





Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Ron364]

Thankyou to Cibachrome, Norm and Greg for the type of replies that I was seeking initially from an engineering forum.

Greg, GM invested huge dollars in seeking a solution, as did the others, way more than you suggested and still came up short. I suspect that they ran out of ideas rather than money. Seems the white flag solution using TC wins in the end! I bet future experimental models will have engine dampers and horizontal rear hub dampers, if they have read this forum that is. Pity the rubber prop didn't find its way throughout GM.

Ciba, I will try to find, borrow or steal different offset wheels and see what happens. Not much clearance to the bodywork but maybe a little will be enough to help. I am keen on Michelin Pilot Sport 4 but they seem to love the low 30/35 series profile that pleases the looks dept but not the usefulness dept. The weight of the huge wheels and tyres these days is massive (mine are 26kg). GOPro vision of a wheel tramping is scary as it wobbles around with great energy. The American fascination with drag racing means that little can be achieved with road going tyre options. I was banned from a web site simply because I pointed out to PeddersUSA that their poly joints were inappropriate due to geometry squish issues. Rule apparently is to never critise a forum sponsor, even if they sold inappropriate "bushes" to an American market that seldom questions authority. PeddersUSA went bankrupt soon after!

Norm, thankyou for your input. Yes likely that GM, BMW, Merc etc used these tools but all reached the same conclusion; to use electronics. Hotup offshoots like Aussie HSV, Harrop and Walkinshaw are good at bolting on extra horsepower like superchargers and soon found the tramp issue in spades. They simply used the "tune the TC" solution hence making the extra power of little benefit. Driving one of their rocket ships, I remember getting wheelspin at 160 kph on dry pavement, a scary experience. Their high HP mods mean the tweaked TC now eliminates wheelspin at a very early stage, and thankgoodness really for these cars are missiles.

Conclusions: don't believe the advertising hype from manufacturers or from (most) vested interest aftermarket. IRS has indeed great advantages over live rear ends, but the downside seems to be you must deal with tramp issues. I am aware that you could always manually lift the throttle, as if that will ever happen! R

 

[GregLocock]

"GM invested huge dollars in seeking a solution," cite please?

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[GregLocock]

When I worked with GM products we used to come up with solutions to the problems we found, and the respective GM division would say, "Oh we know that but we can't afford to put it into the car". Now that may be true, or it may be CMA. A RUT prop is an expensive proposition and at $200 would be a deal breaker for many programs I have worked in, especially if an electronic fix is available.

Holden in particular are not famous for over-engineering their cars.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[BUGGAR]

Info to follow. I will attempt to attach photos first.

 

[BUGGAR]

Here's the other end.

 

[BUGGAR]

I’ve read the entire thread and guess I have little to add. I will be testing my setup for torsional stiffness at the differential pinion with the axles locked – maybe this info will be useful as it will represent a solidly mounted differential carrier with stock (and later poly?) carrier bushings.

I’m putting a 2011 Gen 5 Camaro (Zeta Platform) rear suspension in a ladder chassis. I have the differential carrier installed and ready for load testing. Load testing will consist of locking the axle shafts and applying torque to the pinion flange to load all the connections and parts. The load testing will include measuring the torsional stiffness, which will be collective of pinion axis, and axle shaft axis torsional stiffness. I will use this as a baseline for future carrier bushing replacement. I have installed the carrier mounts rigidly in the chassis but have retained the stock GM carrier bushings for now. I have thus eliminated the stock GM cradle which attaches the entire rear suspension and differential carrier to the structure of the car through four large bushings. These bushings are huge (3 ½ inch diameter) and that has to provide significant deformation for the whole rear end (and likely present secondary vibrations between the differential carrier and suspension as separate flexible elements?). These bushings are designed for lateral as well as vertical compliance. The stock cradle itself appears to be an effectively rigid shell structure and I wouldn’t expect much deflection from it. But the bushings? Also, the Zeta platform has those “L” shaped upper control arms bushed to provide fore-aft compliance at the top of the upright (hub carrier?). The geometry of the lower attachment of the control arms to the upright is not physically describable – you must first thoroughly understand astronomy. The stock driveshaft has Giubo joints at each end and a single universal joint in the center. Enough said about that!

Have you measured the collective torsional rigidity of your entire assembly? Has anyone done this?

Bob

 

[GregLocock]

You should see quite an improvement in secondary ride and handling when you replace the subframe bushes with solid mounts. Road noise and diff whine will get worse. You may see a reduction in linear range understeer, if they have used the compliance of those bushes correctly. " stock cradle itself appears to be an effectively rigid shell structure " - no such thing!

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[BUGGAR]

In rechecking my structural design for the differential carrier mounts, I note that the load on each of the three carrier mounts is different since there is rotation about two axis. Is there any kind of torsional vibration possible from this? I had ignored this, however, if individual axle torsional stiffness is relevant, could two axis sympathetic torsional vibration occur?? I can’t find any references on this. I'm out of my league.

Bob

 

[GregLocock]

Ah, you mean that the torque in the propshaft is 1000 Nm, and in the axle is that times the axle ratio?

What will happen is that the diff nose will not just pitch in response to torque, but the it will also roll. That may explain why people put the third diff mount to one side of the pinion. The net effect is to put some angle into each of the inboard CV joints on the halfshafts, and the rear Gibo. They are all designed to cope with that and are effectively true CV joints so you don't see much vibration from them.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Ron364]

Buggar, well done, nice to know I’m not Robinson Crusoe after all. I have solid cradle mounts and found that of course movement in the cradle is zero. I too rethink the cradle is over designed and did contemplate (for 30 seconds) making up a complete rear tubular cradle mainly to mount the upper and lower “wishbone” arms parallel to the vehicle centreline to avoid any joint squish with spherical joints, stiffer rubbers or poly. But not done so far. The only weak point that I can see is the mounting for the lower trailing arm, GoPro video does show that this bracket is suspect as it seems to vibrate. However, by far the biggest movement is the diff itself so I applaud your work. As previously stated, google “BMW diff lockout” and you will find that using solid diff mounts, even with all else the standard flex rubber, seems to fix the problem in the very similar BMW setup. Note also that 2017/18 M3s also now use a solid cradle mount but stick to flexible diff mounts and electronic TC or “launch control” or similar advertising hype. Just be wary though, the TCs don’t always switch off, and I seem to be the only one who tests on wet roads where tramp is easily the most severe.

I also have poly joints in the diff; a suspect material as rebound in poly is very odd because displacements are better controlled but frequency is higher. My thoughts now are that a rubber snubber or damper is worth experimenting with, given that my engine damper was very successful. Be aware that GM tried this and snapped the alloy diff housing. Buggar, the jig looks good and I feel that a damper is needed though. If I can figure out how to attach GoPro vision to this forum, it makes for a fun movie. File type needs a special viewer though. Greg, solid cradle mounts improve the handling as expected as the result is that the rear takes a “set” quicker entering a corner. There is not a lot of extra noise, but yes there is a bit more. I am thinking that Cibachrome’s view that tyre (tire for some of you!) sidewall stiffness experiments may help. 26 kg wheel/tyre at resonance is a big worry too. Funny that circuit racing cars all use high profiles like 50; these silly 30 series profiles for the road must come from the non engineers in the team. Rollcentre is close to where the driver’s hips are and too hard to change so far to get more squat. (Camber change works against you though as it squats)

 

[cibachrome]

Instead of loading the pinion nose with a large moment to measure the roll and pitch stiffness, I suggest you use the proper driveshaft set in a fixtured position to represent the transmission tailshaft angle and height. Under full diveline torque load and suspension jounce, the U-Joint angle is going to tip the pinion nose to another sideview angle. I suggest you make your setup such that the driveline is aligned straight under this load instead of just layout position. Plenty of videos on this. You should know that 'some' racers run their motors opposite direction to make this problem disappear. Its a fantastic little know 'secret', but there are a few unsderground 'reverse' motor accessory shops with zippered lips. It all started with water pumps and took off from there.

BTW: The Camaro's have about 4% rear roll understeer (tad less on the SS) and about .08 deg/1000N lateral force understeer. Corvette's values are about half this, but the Mz oversteer is twice (as you could imagine).

 

[Ron364]

The attach photo link shows a simple diff bracket to suit a BMW. The link to a description is

https://bmw.spoolstreet.com/threads/clark-performance-fabrication-diff-lockdown-install-on-dct.2478/

Not sure of long term use, but hey, don’t knock it as there are plenty of BMW owners that say it works. Bob, yes the welders solution, but certainly worth a try.