Toe / SAT

[Overlap77]

Hi
i'm reading this page

https://www.eng-tips.com/viewthread.cfm?qid=432856

but I can't find what the SAT in toe/SAT stands for.
Could someone please help so I can fully understand the points made in the discussion
thanks

 

[Overlap77]

sorry i found it - in Bundorf analysis on wikipedia

"SAT compliance steer. The aligning torque directly twists the wheel on the compliances in the suspension, generating a steer angle. "

 

[Overlap77]

ok i will try and ask a question without answering myself that relates to this topic.
Greg was the BMW you mentioned an E36? I remember driving one for the first time, and wondering why the steering felt so "dead", could this be the reason?
thanks

 

[GregLocock]

SAT is self aligning torque, ie the torque around the vertical axis. Often we think of it as the product of the lateral force*pneumatic trail.

It's a funny name.

That BMW would have been a late 80s 5 series. I can't remember if the X5 had a similar isolator, I don't think I was ever allowed to strip it down!



Cheers

Greg Locock


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

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

 

[Overlap77]

thanks Greg - if I understand the point Cibachrome was making about SAT, that modern cars have been "stiffened" in so many areas that SAT is used to try and dial in the understeer? I thought SAT was something that was describing tyre behaviour rather than slop in the steering system.

 

[GregLocock]

I doubt that's what cibachrome meant. Slop, or lash, is something you'd never want in a steering system, whereas compliance may have its uses tho I struggle to think of an example where I have willingly put compliance in. Even though steering column torsional compliance counts as understeer in some ways of defining it, the sacrifice in steering feel ( a rather nebulous subjective term) has not seemed worthwhile.

Cheers

Greg Locock


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

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

 

[Overlap77]

You're quite right. It's all to easy to throw in the wrong terminology without thinking.

to quote " The rather LARGE values for vehicles on these plots can surprisingly come from vehicles which might be puffed up as world class state of the art right brained fantasies, but in fact are almost entirely saved by way oversized tires which have very high cornering stiffness used to minimize the overall impact of the required soggy steering system."

Does this apply to all modern vehicles or mainly SUV Type? The soggy steering system is required due to advances in shell stiffness?

I'm trying to understand Magic Formula tyre charts

The Mz (same as SAT?) curve - i can see it hits zero as the limit of grip is reached but the peaks of Mz seem small in comparison to lateral force. Does a bias ply tyre have higher Mz plots? What does high Mz vs low Mz correlate to from the driving seat?

thanks

 

[GregLocock]

Oversimplifying: In a manual steering car the steering wheel torque, SWT is proportional to the total moment about the steer axis of the tire. That comprises two parts, Mz, or SAT (yes the are the same), and the mechanical trail*the lateral force (Fy). A slightly less weird way of thinking about is to define Mz=pneumatic trail*lateral force, then the torque around the steering axis is Fy*(mechtrail+pneutrail)

If you had no mech trail then as the tire hits peak Fy you wouldn't feel any SWT. That's why the steering goes light towards the limit. If you have loads of mech trail then SWT is proportional to Fy, which masks the peak of Fy which is what you want to sense.

So the optimum mech trail is somewhere between 0 and a lot, which doesn't help much, but typically 20-30mm is used, roughly the same as the pneumatic trail as it turns out.



Cheers

Greg Locock


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

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

 

[cibachrome]

Take a look at this 'executive summary', all from measurements of chassis and tire properties. The red bar is the Mz steer component of the Cornering Compliance summary. Some of this is philosophy, steering isolation (as in broken fingers, vibration, etc.) and some is desperation. In power assisted steering, the shape of the assist function (as in hydraulic valve profile, pump flow rate, tie rod load gradient, or electric current flow) are the tuned functions that load up the compliance elements. Because of weight distributions, suspension DNA, load conditions and spare tire considerations the elements of the Bundorf recipe can be all over the place. Sometimes they have to resort to different rear tires, wheels and pressures to make the vehicle saleable. Bad for tire rotation schemes, flat tire problems and wheelhouse size impacting trunk space. (OK, boot volume).

 

[cibachrome]

As for your other question(s) Mz peaks before Fy peak so drivers often fail to use the max capability of the vehicle unless you fix this as Greg mentioned.

Bias tires have always had higher Mz and camber output, hence the great apprehension when putting radials on a vehicle that had bias/belted tires. Funny, burn down a 'Racing Radial' and note how little radial is there. Racers like/want/need the camber gain to stay competitive. And the 'feel' and the Mx contribution or (lack of).

As for the Bimmer, here's a clue as to how they work with impedance matching via the steering isolator's hysteresis properties. It's like a viscosity not just stiffness. Stiffness testing on the components is done at several input velocities to get all the descriptive parameters. I forget the hysteresis metric, but the others are 5 Series Bimmer class steering shaft properties at about .25g. If you believe that stiffness is the only specification to give to a supplier, its remarkable how much a 'softer' shaft is preferred, to the great dismay of geeky handling ex-spurts.

 

[Overlap77]

that is fascinating, thank you. I'm busy today but I'm going to study that and try and say something intelligent !

 

[GregLocock]

I'm wiki, ciba is SAE paper!

Cheers

Greg Locock


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

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

 

[Overlap77]

So as I look at the data for the FWD Saab with Strut front and a twist beam rear and compare to the Jaguar, with Double control arm and multilink rear, it's hard to discern much of a difference. I'd often wondered how people could get so misty eyed about certain cars, yet when you drive them they aren't as special as the review would have you believe.

Would it be fair to say that the real challenge is creating something "special" when all the numbers point to a vehicle being very similar to its peers. For example how could anyone tell from the steer result data that a 911 would make so many journalists and owners go wild with delight?

Or do I just need to see the graphs on a much bigger screen!

The intelligent comment will have to wait for another day.

 

[GregLocock]

Defining objective metrics to represent what a skilled observer prefers is a huge body of mostly proprietary knowledge. FWIW linear range understeer (defined in a certain way) certainly is one metric, and it feeds into several of the others. It would be very hard to look at the K&C type plots from two cars and guess which one was nicer to drive (defined somehow or other) if they are both reasonably competent.

Cheers

Greg Locock


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

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

 

[cibachrome]

The graphs only indicate approximate steady state control allocations and await actual steering sensitivity (g/100 deg), and steering torque gain (Nm/g) as well as how they are balanced (Steering Work Sensitivity is basically their ratio), with a nominal value having been identified for 3 kinds of drivers: average, sporty and extreme. Manual steering vehicle tends to be left out in the cold because steering effort is not assisted, hence arm strength is required for parking the stove.

Pay attention to the DF and DR terms (AKA Cornering Compliances). Transient response (response time, settling time and damping factors are related to the sum of the absolute cornering compliances over the product. So, having low DR is a key ingredient with just enough DF to produce a decent K (understeer). You need some K to slow down the rate of change of steering gain with speed and to allow for payload changes (passengers, delta fuel load and luggage while maintaining a legal level of stability. One other need for K comes into play when you replace worn out tires. (95% of customers go cheapest, not best, manufacturer mandated or even major brand name. Just what's on sale.) Quite a few date bait vehicles have 'Summer' tires (good dry grip, awful snow traction and lousy fuel economy) that are supposed to be swapped out when the snow flys. But they don't, as you can tell from the Youtube videos of winter spins, crashes and other comedy channel displays. Some folks just never change tires. Lucky ones get pulled over by LEOs when the sparks from their tire's steel belts aggravate cars following them. Then there are the ones who love the ride quality of their favorite sleds with 15-20 psi in the tires. Did I mention that 5 year old high pressure spare tire that's leaked down to atmospheric pressure, having never been checked (Hey, what the hell kind of tire is that in by boot. ???).

Finally, great handling requires extra cost: premium tires, lower fuel economy, premium materials, extra care in engineering and development time, and extra attention in assembly (for quite a few reasons I can think of).

But the big bars in those plots (I'm delighted to have found a way to put several pics on the one allowed forum image) are tires. Big (lotsa load reserve) and stiff (extra baloney parts in the carcass) tires make smaller blue bars and their sum/difference gets you the best shit eating grin when driving off the sales lot. The other colors are residuals from wheel bearings, control arm, strut and steering gear mounts, packaging demands, comfort, cost, shared components from other platforms and lousy theories about what the world needs next from their 'New and Improved' model. Just like laundry soap !

 

[cibachrome]

FWIW.2:
If you can run an ISO Frequency Response test procedure (finite steer pulse or swept sine steer) to get understeer(deg/g), lateral acceleration response time (sec), peak roll gain (deg/g) (not the steady state flavor) and steering sensitivity(g/100deg at 100 kph), you can make a delightful listing of all your FR test results from: K * LART^2 * phi_peak* exp(((ss-1)/.8)^2 and look what/who marches into the top zone. It says: keep your understeer minimum while getting really fast response times, good roll damping and nominal steering gain. Efforts to clamp down steady state roll gradient are a lost cause because the peak roll jumps up, requiring unlivable roll damping (from RIDE dampers that can rupture vital body parts and release unwanted bodily fluids). And even break parts when necessary ! This rule by thumbs pointed upward is really just common sense...

 

[Overlap77]

thanks.

I don't understand this
"Efforts to clamp down steady state roll gradient are a lost cause because the peak roll jumps up, requiring unlivable roll damping (from RIDE dampers that can rupture vital body parts and release unwanted bodily fluids). And even break parts when necessary ! This rule by thumbs pointed upward is really just common sense... "

When you say clamp down on steady state roll, would that be through the use of anti-roll bars? If so how does clamping down on the roll gradient cause greater peak roll? I didn't think dampers were generally used to control anything but initial response? Although a Dutch manfacturer says they have developed a system that controls roll enabling softer springs and without affecting comfort. I imagine this is some kind of frequency specific damping.

 

[cibachrome]

Greg: If you buy into the K-Tay-Pk_Phi_SS neural net findings from customer clinic training, the 'Top' 50 cars from a by-gone era were as follows. The details are usually interesting because they show good a 'good' vehicle can be dropped out of sight for just a 1 metric penalty. Response time/bsndwidth is obvious the dominant player.

2006 PORSCHE CAYMAN S No. 1
2006 PORSCHE CAYMAN S No. 2
2003 ACURA NSX
2000 HONDA S2000
2006 BMW M5 No. 1
2006 BMW M5 No. 2
2005 TOYOTA TACOMA X-RUNNER
1999 DODGE VIPER GTS
2006 BMW 650I
2003 INFINITI G35
2006 HONDA S2000
2005 CADILLAC STS No. 1
2005 CADILLAC STS No. 2
2006 BMW 530
2000 SUBARU EURO IMPREZA (WRX)
2002 MITSUBISHI LANCER EVOL VII
2004 OPEL ASTRA
1993 TOYOTA SUPRA
1999 BMW M3
2002 MITSUBISHI LANCER EVOL VII
2005 CADILLAC SRX
2003 NISSAN 350Z TRACK
1994 ACURA NSX
2006 MERCEDES CLS500
2004 HONDA S2000
2006 LEXUS GS430
2000 PORSCHE 911 CARRERA
2002 MITSUBISHI LANCER EVOL VII
2004 AUDI A8
2007 ACURA RDX
2007 HOLDEN COMMODORE SS-V
2004 BMW X3
2004 BMW 745I
1999 FORD MUSTANG SVT COBRA
2006 HONDA CIVIC SI
1996 MG F
2006 MERCEDES ML500
2004 BMW X3
1999 HONDA CIVIC SI
2007 ACURA MDX NO. X879089|
2007 MAZDA SPEED 3
2001 JAGUAR XK8
2006 LEXUS IS350
1998 PORSCHE 911 CARRERA
2003 PORSCHE 911 TURBO X50
2004 SUBARU IMPREZA WRX STI
1999 LOTUS ELISE
1999 BMW M COUPE
2006 INFINITI M45
1998 CHEVROLET CORVETTE

 

[Overlap77]

This forum's much more interesting than car magazines, and yet, here is compelling "evidence" for Porsche actually being technically superior, rather than the journos constantly saying they are the best, but never giving technical reasons why.

I notice the MGF is in there. When these came out they were very highly regarded but over the years the usual bias has made almost everyone think the MX5 is better. Was the MX5 in there ? Although the MGF was stiff for a convertible of the time you could tell the difference with roof up or down in terms of scuttle shake.

Stand out contradictions would seem to be the Tacoma and the X3 though? No objective assessment from behind the wheel would put them in that list. I guess it shows a "magic formula" is like the holy grail..

Any supercars in the results?
I wonder how the latest Civic Type R would fare.