Total Lateral Load Transfer Distribution "Rule of Thumb" 3

[mark512]

I've come across this rule of thumb in a few places that the total lateral load transfer distribution (TLLTD) with respect to the front axle should be around 5% greater than the static weight distribution with respect to the front axle. What's unclear to me is if this percentage is multiplied in or added?

For example, a car with 35% of the weight on the front axle:
35% inflated by 5% is 36.8%
35% + 5% = 40% (which is 35% inflated by 14%)

Thanks in advance!

 

[BrianPetersen]

This sounds like a rule of thumb that perhaps ought to be forgotten about. Or maybe it applies in a narrow range of circumstances - but then, a stopped clock shows the right time twice a day.

Porsche 911 (rear weight biased) in heavy cornering are known to lift up the inside front wheel.

VW Golf (particularly Mk1 and Mk2 - front wheel drive) in heavy cornering are known to lift up the inside rear wheel. This suggests that the rear is pulling more than its weight in terms of load transfer distribution.

Front-wheel-drive cars already understeer heavily without trying to make the front handle too much of the load transfer.

 

[Tmoose]

Hi Mark,

Some info about that thumb rule's context would interest me.

thanks,

Dan T

 

[mark512]

From here:

http://www.optimumg.com/docs/Springs&Dampers_Tech_Tip_2.pdf

At OptimumG, we call the roll gradient distribution the Magic Number (Milliken calls it Total Lateral Load Transfer Distribution). The Magic Number is expressed as the percentage of the roll gradient taken by the front suspension of the car. As a baseline, use 5% higher Magic Number than the static front weight distribution.

Though, it seems like my original question might be answered here:

http://performancetrends.com/text-files/CTA Suspension Report.txt

Comments:
Your FLLD of 43.4% is much lower than typically used.
A suggested 'starting point' for FLLD (Front Lateral Load Distribution) is 55%.
This is based on adding 5% to this vehicle's 50% Front Weight Distribution.

The "magic number" seems to be used in a few weight transfer worksheets:
optimumG:

http://members.iinet.net.au/~bushfam/jason/Magic Number SI V1.xls

Smithees Racetech:

http://www.locost7.info/files/suspension/Suspension_spreadsheet.xls

 

[BrianPetersen]

"Rules of thumb" are OK as long as you don't violate the underlying assumptions ... which requires knowing something about what those assumptions are.

I am going to GUESS that the assumption here is being made that one is dealing with a front engine rear axle drive car with somewhere in the vicinity of 50/50 front/rear weight distribution. (About that stopped clock being right twice a day ...) Granted, lots of sports cars and sporty cars are like that.

It's not uncommon to want the axle opposite the drive end to take more of the lateral load transfer in the interest of keeping the weight on the drive wheels more evenly distributed - to minimize wheelspin when exiting a corner. In the case of rear axle drive, it helps with the understeer budget as well.

I will just about guarantee that trucks will violate this assumption. They need lots and lots of roll stiffness on the load-carrying wheels.

Front wheel drive will certainly violate this assumption unless you want an understeering pig with no grip coming out of corners.

 

[cibachrome]

This "Rule of Thumb" does not originate from design space, it shows up from the architecture of the car and the targeted performance features. How much tire reserve (rated load vs. usage load), same tire size all 4 wheels ?, engine power, number of motor types for each model, the number of goofy (abnormal) suspension design factors (like roll steer front and/or rear), ride balance, passenger usage vs. capacity, AWD vs. RWD, and a few more, all play into the final frosted cake.

"Average" tires (10% - 20%) reserve with decent ride rates/wheel travel, 2+2 passenger load rating, targeted tire section/mass ratio, average tire pressure, rear drive, small gas tank, etc. may wind up with 5% TLLTD when all the ride and handling development is finished. This would be done with a decent sized front bar with good efficiency (no sewer pipes), and a small rear bar to finalize the roll gradient. There's some camber by roll coupled with tire overturning moment output to factor in, too. (Anybody ever heard of snap-through?

Nobody's mentioned the 'other' part of load transfer distribution because all the handwavers, book thumpers and academia have no clue about the data or ways to figure it, and that is what the DTLLTD is all about. This is the effect of Dynamic load transfer from your shock absorbers/dampers. No need to go into complex valving profiles because the roll frequency velocity inputs are low compared to a max pothole strike. These are asymmetric to, BTW in case you forgot. And the DTLLTD may/will be negative.

For a Level III or Level IV handling car (Ride & Isolation, what's that?), tire reserve may be 40%. TLLTD could be negative because your large low section/mass tires actually like load and amplify it (a bigger front bar adds mucho oversteer). Once again, pressure, brand and construction details are strong players in this recipe. This means the DTLLTD has to be conditioned for this too.

Of cars mentioned, they often have goofy suspension parameters (front or rear) to 'fix' a low understeer condition (often in spite of using different front and rear tire sizes and pressures). Knowledgeable geeks would know that tires stop listening to steer (as in roll steer for example) when the cars get pushed so they run out of stability and get a reputation for "speen outs". Needless to say that they also don't roll a lot, hence the gigantic steer by roll and camber by roll K&C numbers.

So, you need all the ingredients to bake the cake, not just the rule for 1 cup of sugar and a teaspoon of salt. Frosting covers up a lot terrible cakes but bakers know !

 

[mark512]

How much tire reserve (rated load vs. usage load)

I haven't come across "tire reserve" before but it does make a lot of sense. When you're talking about these percentages, what do you take as the rated and usage loads? My assumption would be that the rated load is the value indicated by the load index letter on the tire, and that the usage load would be the worst case load during some high-G maneuver (like the load on the front outside tire during simultaneous braking and cornering at the limit...)

 

[cibachrome]

Not quite. The rated load is the sidewall reading at the sidewall pressure. The usage load is the max static load any wheel will see (as in GVW (all passengers + full gas tank + labeled payload amount). These days, passenger weights are a bit light (based on Millenials, instead of the more appropriate Snowflake Tide Pod ingester weights). Here's a look at the real world. I'll let you figure out what tires go with what cars. Tires with high load sensitivity (they don't mind increasing loads) go hand in hand with FWD for reasons you already have heard about. In your case (I'm guessing), the wider wheel rim is also a money maker as long as it fits your brake package, has friendly wheelhouse contact and your steering system has enough grunt to keep up. And, rim width difference front to rear works nice too, as long as you can deal with the spare tire issue (h.p. spare, plug kit, air pump, Slime, Coates 4040 in the trailer, or a wheel sparkling all the way home).

 

[GregLocock]

There is an SAE paper called magic numbers 911921. It rounded up many 'typical' rules of thumb for suspension tuning. I agree with the above, it is somewhat relevant for front engined rear wheel drives. FWD cars are actually more difficult to get right, in my opinion, this may even be why the original Mini was somewhat skittish (hilarious as it was).

The practical limit is that your subframe is only so stiff, so above a certain diameter increasing the sta bar size has little effect on roll stiffness, and a larger effect on weight distribution. The other thing I see a lot of is cars that use a lot of roll steer to get linear understeer do not get more understeer if you add front bar, the extra roll stiffness prevents the roll steer from working, so understeer changes little.

In road cars there is a good argument that roll gain is not a very good metric for the subjective impression of roll, I wish we had a better one. I suspect damping has much to do with it.



Cheers

Greg Locock


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

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

 

[cibachrome]

Try looking at the roll peak to steady state ratio obtained from a finite pulse width or chirp frequency response test (roll angle by lateral acceleration). You will find a distinct set of statistical buckets that cluster by continent. Japanese sleds will be 1.1 to 1.4, Euro cars will be 1.5 to 2.0. North American designs can be 2.0 to 4.0, although this changed quite a bit once the trends were identified. Ride focus and available wheel travel can be a players here as well as road PSD and even availability and usage of (high cost) premium dampers.

This can get nasty when lightly damped (in roll) vehicles with yaw velocity and sideslip coupling occurs. Roll factors are mostly independent of speed. Not so with yawrate and sideslip. When they hookup and crossover, you have a problem. Roll understeer becomes roll oversteer for example. (and why some VERY FAST cars use roll oversteer). This also a roll center (axis) determinator (is that a word ??), since ride damping is the selector and total roll inertia and stiffness are step-children of the platform architecture and marketing calls.

 

[GregLocock]

Peak/SS is a nice measure, yes I'll have a chat to the ride guys.

Cheers

Greg Locock


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

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

 

[mark512]

Not quite. The rated load is the sidewall reading at the sidewall pressure. The usage load is the max static load any wheel will see.

That makes sense, as a few tire manufacturers offer calculations for determining the required tire load index, and all of them just boil down to the static weight on the tire.

There is an SAE paper called magic numbers 911921. It rounded up many 'typical' rules of thumb for suspension tuning. I agree with the above, it is somewhat relevant for front engined rear wheel drives. FWD cars are actually more difficult to get right, in my opinion, this may even be why the original Mini was somewhat skittish (hilarious as it was).

I'm going to take a look at that paper. And the whole applicability to front engine / rear wheel drive makes sense, as a lot of these things seem to come from books written in the 70s and 80s by American authors for people interested in building a racing car, so an FR configuration sounds like what the vast majority of their readers would be dealing with.

Try looking at the roll peak to steady state ratio obtained from a finite pulse width or chirp frequency response test (roll angle by lateral acceleration). You will find a distinct set of statistical buckets that cluster by continent. Japanese sleds will be 1.1 to 1.4, Euro cars will be 1.5 to 2.0. North American designs can be 2.0 to 4.0, although this changed quite a bit once the trends were identified.

Very interesting, but what units are you using to get these ratios?

 

[cibachrome]

Come on, these are RATIOS of peak value to the zero+ frequency (Steady State) value obtained from a magnitude Bode Form transfer function computed from road test measurements. The P2ss ratio is interchangeable with the damping factor 'zeta' used in control theory. So, for this example, the 2.08 [deg/g per deg/g] roll gain P2ss ratio from Bode Form analysis is equivalent to a zeta of 0.246, although this is more commonly inferred to be from a time response signal. This zeta comes from the denominator of the computed transfer function.

 

[mark512]

Got it, I was reading roll by lateral acceleration and was thinking it was some ratio of the two, but yes, peak vs. steady state of the transfer function of lateral acceleration to roll. Makes sense.

...though it's interesting that the Japanese cars have the lower damping and the Europeans are in the middle - I would have thought it was the other way around, however, given that my first car was an older Buick, I can definitely relate to the American numbers!

 

[cibachrome]

The Japanese have HIGHER damping ratios according to the stats (low P2ss ratios), although the architecture (wheelbase and track) motor weight, roll axis height, roll gain and a few other obtuse factors play into this. The synthesis of vehicle subsystems by some manufacturers is driven by just a few vehicle and marketing requirements and leads to these performance buckets, and it becomes a signature of the brands. For others, hap-hazard copycat and monkey wrenched 'designs' give you a different vehicle every year because the tweaking never stops. "Refined", "Improved", "Updated" are the catchy buzzwords for the uninformed.

 

[mark512]

There is an SAE paper called magic numbers 911921. It rounded up many 'typical' rules of thumb for suspension tuning. I agree with the above, it is somewhat relevant for front engined rear wheel drives. FWD cars are actually more difficult to get right, in my opinion, this may even be why the original Mini was somewhat skittish (hilarious as it was).

I did end up getting a copy of that paper, it was a pretty good read, and really did only deal with front engine rear wheel drive (or at least only considered cars with double wishbones up front and a live rear.)

The Japanese have HIGHER damping ratios according to the stats (low P2ss ratios)

I think it clicked now, is P2ss analogous/synonymous with Q-factor? With the relation Q = 1/(2*zeta), the numbers jive with the example you gave.

This can get nasty when lightly damped (in roll) vehicles with yaw velocity and sideslip coupling occurs. Roll factors are mostly independent of speed. Not so with yawrate and sideslip. When they hookup and crossover, you have a problem. Roll understeer becomes roll oversteer for example. (and why some VERY FAST cars use roll oversteer). This also a roll center (axis) determinator (is that a word ??)

I think I get what you're saying, and it makes sense, but to confirm: light roll damping + roll/yaw coupling (i.e. very inclined roll axis) = potentially nasty?

 

[cibachrome]

By Jove, he's got it ! Now you should consider yourself a Global Vehicle Dynamics Champion, write up a mag/rag article, publish a book, join a Racecar team as Chief Engineering Officer and get a discount on parts at AutoZone.

Its usually high levels of rear roll understeer that raises the eyebrows and soils your shorts in cars with low roll damping and 'performance' values for yaw velocity and sideslip natural frequencies. They try to bite down the steady state roll which helps at legal highway speed, but going faster (140+ kph) can get you into big trouble. Cars with 'large' high performance tires with lots of tire reserve (remember what they like ?) are the bad actors. Porsche 911 / Carrera stats come to mind. 13+ deg/deg rear roll understeer ? For what ? And their reputation precedes them.

The way we test for this susceptability is to run ISO-8726 Frequency Response tests at multiple speeds, starting at a 'safe' constant speed (say 80 kph), then a few 20 kph speed increments. Plotting the yawrate frequencies and damping levels vs. speed tells you what the Grim Reaper has in mind for you at the advertised max speed capability.

 

[BrianPetersen]

I'm not so good with the formulas and numbers but cibachrome's explanation of the theory explains an incident years back in my friend Al's early-eighties Dodge van ... (the old rear drive 225 Slant Six B van, not the minivan)

To see what would happen, "just because", he jerked the steering wheel back and forth a couple of times to get it rocking side to side while travelling at highway speed. In retrospect ... Maybe not such a good idea, but it seemed funny at the time.

After he stopped the steering wheel input, the van kept on rocking side to side, quite violently! If anything, the amplitude increased, probably to the point of hitting suspension travel limits! That ended when a tire blew. Amazingly, the van stayed on the road despite the blown tire, but he's lucky it didn't end in an off-road excursion and roll-over. The tire that blew, tossed its tread up against the fuel filler neck, partially flattening it. Of course, this wasn't realized until the next time it needed fuel ... and it could only be filled at a trickle!

It's pretty common for truckish vehicles to have leaf spring rear suspension with the leaves oriented in such a way that it pulls the axle forward with suspension compression (rear leaf-spring eye higher than the front one) ... that's roll understeer.

It's pretty common for truckish vehicles to have the rear shocks mounted inboard of the leaf springs and frame rails ... which makes them much less effective in roll than in two-wheel bump. Not much roll damping.

In this particular case, it probably didn't help matters that the vehicle was a tired old heap, and I'm sure the ball-joints, bushings, steering linkage etc had all seen better days, and I doubt if the tires were anything to write home about, and "handling" was the furthest thing from the mind of its designers even when it was new. High center of gravity, too.

I know that roll understeer is built into a lot of suspension designs but it seems like while a little bit is perhaps OK, a demand for a lot of it is indicative of some other problem.

Some front-drive vehicles seem to have the roll understeer (a.k.a. "intentional bump steer") built into the front suspension as opposed to the rear, which is a twist-beam axle as often as not.

 

[cibachrome]

Roll understeer is built into many suspensions because it's a factor in two aspects of vehicle handling:

It increase the vehicle's yawrate and sideslip natural frequencies which lowers the response times. This is an easier way to do it rather than using rear tires with higher cornering stiffness (different load sensitivity, different rim width, different pressure, different size, etc). That's a packaging, cost, part numbers, tire rotation, spare tire usage headache.
The other 'feature' is the reduction of rear cornering compliance loss as you add rear payload (passengers, trailer hitch load, full beer kegs in the trunk). This reduces the increase in steering gain and lowers the increase in response times (that description looks like nonsense on paper doesn't it ?) that would arise from mass and mass distribution change. In fact, it is possible to have a vehicle whose understeer/steering gain and response times are about the same no matter what the load condition is. That makes for a driver's warm and fuzzy feeling. In trucks, this would seem ideal, except now they add Level Control with rear air shocks to put all this in the tank.

With twist axles, the amount of roll steer is defined by the difference in height between the control arm bushings and the shear center of the cross beam. There are V beams, U beams, I beams, O beams and Phaser beams. The roll steer arises from the warping of the beam flanges (play with a rolled piece of paper). The warping action causes mayhem from a manufacturing standpoint as it works the attachment to control arm welds very hard. If anybody remembers what a wire clothes hanger does when you bend it a few times, you will get the picture. Terry Satchell wrote an SAE paper on all of this. And, using Phaser beams does not require you to drop out of warp (as is required for photon torpedo use.

 

[GregLocock]

Satchell's very useful paper is at

https://www.sae.org/publications/technical-papers/content/810420/

Cheers

Greg Locock


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

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