Why Mid-engine 4

[BUGGAR]

So, what’s up with new trend towards mid-engine performance cars? There’s mumbo jumbo about weight distribution. But no one has mentioned Polar Moment of Inertia. I’ve been watching films of race cars spinning out. Some recover quickly, some do not. Comments?

 

[BUGGAR]

Going back to the significance of a mid-engine vs front engine for a lower vehicular polar moment of inertia:

MOI of 4 wheel/tire assy’s x 150-lb ea. x 5’^2 = 15,000-lb ft^2

Front engine MOI from 2-ft offset = 800-lb x 2’^2 = 3,200-lb ft^2 - Insignificant in comparison, therefore engine placement does not affect the MOI that much.

Bob

 

[GregLocock]

Well, that's the opposite of my conclusion. I may be able to get C7 and C8 inertias.

Cheers

Greg Locock


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

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

 

[BrianPetersen]

What axis are you calculating this MoI around??

 

[NormPeterson]

Buggar - wouldn't it be better to sum the inertias of the wheels, powertrain, and everything else lumped as a rectangular mass about the CG . . . and find the difference between the totals?


Norm

 

[BUGGAR]

Norm, yes it would but I thought I'd try a ruff 'n dirty first.
The MOI was considered midway between wheelbase and track.
Bob

 

[BrianPetersen]

When you steer with the front wheels, at least for an instant until weight transfer and suspension movement and tire compliance start having extremely complex confounding effects, the car yaws around the (un-steered) rear axle. (The front goes sideways and the rear stays put relative to a frame of reference travelling with the car's initial motion.)

 

[GregLocock]

Aye, I suspect the rearward CG has a lot more to do with perceived nimbleness than more complex concepts.

Cheers

Greg Locock


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

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

 

[NormPeterson]

Norm, yes it would but I thought I'd try a ruff 'n dirty first.
The MOI was considered midway between wheelbase and track.
Bob

It's not particularly difficult as a spreadsheet. Though I do think you would need to include a manual goal-seek step to capture the effect of the CG moving as a result of moving a couple of rather heavy items by significant distances.

I'm getting numbers within Greg's range for the total but numbers for the engine plus transmission more consistent with yours.


What Brian and Greg have noted above reads just like a free body diagram of a beam that suddenly goes out of equilibrium due to the appearance of a new force at one of the support locations. I'm not surprised that this is not addressed directly in any of the enthusiast magazines or even the various message boards. Easier to use the dumbbell analogy (no, I'm not going to say it).


Norm

 

[BrianPetersen]

One other small thing. The Corvette has always (since C2, and I'm not sure about C1) been "mid-engine". It has been front-mid up until now, and now it is changing to rear-mid. And the transmission has been at the back since C5. So your starting point for a MoI calculated around the rear axle, or around the CG, was favourable to begin with. The vast majority of the drivetrain has always (since C2) been between the wheels.

The big difference now, I predict, will be what the driver feels. Up to now, the driver has been very close to the rear wheels, and what happened out front was way out front. Now, the driver's head will be close to the middle of the car.

I don't have experience with what that will feel like ... only the opposite. My front wheel drive Fiat Ducato van is as opposite to a Corvette as you can possibly be. Drivetrain hanging out front outside the wheelbase, driver sitting almost on top of the front wheels, what happens out back I have no idea, but being front wheel drive, what's out back is just following along anyhow. When maneuvering at low speed, you certainly get the impression of moving sideways as you steer. It's slow. It understeers. It's not nimble. It's tall. But it hauls my roadracing motorcycle around, and it's my portable garage, and that's the point.

 

[NormPeterson]

Quick back of envelope calc suggests FE to ME is worth 1200 kg m2, which is significant, typical PMOI is 3000-5000 kg m2

Hmmm . . . summing MOIs about the rear axle line makes that 1200 number look much better (I got 1500-ish overall for the model I'm using), without going too far outside that 3000 - 5000 (5700-ish, ± 750-ish). Both of which would of course depend on specifics.

Looking at this from the rear axle line shows a far more significant effect going FE to ME than looking at it from the CG.


Norm

 

[cibachrome]

What matters to the vehicle dynamics of a car or truck is referred to as k' (Kay prime). It is [ the ratio of the Z (yaw) inertia to the vehicle mass -1.0 ]. A value of zero gives you a bar bell. This factor is the player which determines what tire and suspension properties are needed to deliver prescribed Steering Gain, yaw overshoot (yaw damping) and lateral acceleration response time. THESE factors are usually determined by 'competitive assessment' (i.e. what do drivers want or prefer in a designated 'handling emphasized' market slice. A synthesis done by determining the tire and suspension properties necessary to achieve these goals is easily done. What pops out of it is a k' value that is needed to obtain a set of tire properties (lateral force and aligning moment stiffness) needed to meet such goals. Sometimes these goals are unobtainable because there are no such tires which can deliver the specifications.

For example, an equivalent BMW 5 Series on it's 'best' handling/feeling tires (Continentals) can NOT be made from a Cadillac CTS-V because the V8 engine plus it's transmission location can not deliver a k' of -0.1 Instead, it's k' of 0.2 penalizes the needed tire stiffnesses to produce the balanced car. Yes, you can utilize different front and rear tire sizes, but now we are into marketing, part numbers, wheel rims, tire rotation, aftermarket tire availibility, warranty and a few other things some car buyers, dealers (and some LAWYERS) will NOT tolerate.

Here's a simple example:

Lets say you drool over a Kleptomatic Level-III handling car that has nimble handling, great road feel, a wapping max lat and costs less than 70,000 denariuses (denarii ???).

You measure a few and here are it's metrics:

Steering Gain (g/100 deg SWA at 100 kph): 1.80
lateral acceleration response time (sec.) 0.28
Yaw Velocity Peak to Steady Ratio 1.05
[Same as a zeta of 0.69 ]

for k'= 0.2:
Front Cornering Compliance == 4.54 deg/g
Rear Cornering Compliance == 1.83 deg/g
[Understeer == 2.71 ] deg/g
Overall Steering Ratio == 11.80 deg/deg
This much understeer kills your max lat.
The low steer ratio makes for way too high of a steering gain at 250 kph.

Next is k' = 0.0
Front Cornering Compliance == 3.87
Rear Cornering Compliance == 2.01
[Understeer == 1.86 ]
Overall Steering Ratio == 14.40
This is do-able, but where will all that front compliance come from? Roll steer, soggy bushings, high effort steering feel, high caster ???

Then there is k' = -0.02
Front Cornering Compliance == 3.51
Rear Cornering Compliance == 2.21
[Understeer == 1.30 ]
Overall Steering Ratio == 16.82
That 17:1 ratio means its a dog in the parking lot (low speed gain).
The low understeer might make some lawyers nervous in light of the replacement baloney skins some clueless tire dealer will try out.
These low cornering compliances will require some stiff tires, possibly half tread, produce poor isolation from road bimps, be VERY expensive, and result in a gigantic steering gain at max speed. And, the tires can't make up the whole enchilada. You are gunna need some extra compliance(s) (roll steer) this is the Porch Glider solution. Not worth a crap at max lat so Happy Tails to you, Roy Rogers. How about the Lond Rova approach ? Cheewing gum mounted steering gear ?

So, what's in YOUR wallet ?

 

[cibachrome]

BTW: Summing up point masses * distance^2 to get total inertia can yield a 40 - 50% error because you also need to add in the inertia(s) of the 'point' masses. ( Remember your parallel axis theorem ? ) For a dressed engine + transmission, this ain't something to ignore. Cross-products, too. This is the same issue that confronts the designer who is all wound up about where to put the 'roll axis'. The sprung mass has it's own moment of inertia and transferring it to the roll axix has it's own set of requirements and specifications based on resulting roll and yaw rate frequencies. (Remember that one of them is speed dependent and one is not. How did Shakespeare 'splain it ??? : "Never the twain shall meet".

 

[drwebb]

OK, for the folks watching at home- so if there's no technical basis for switching the 'vette from front (front-mid) to mid (rear-mid), then we conclude it is purely marketing-driven, perhaps to attract younger buyers who are cross-shopping (or will only consider) the much pricier import competition? Seems an odd strategy, but in an environment where the experts conclude branding it a 'Mustang' is the best strategy to sell a full electric crossover, I'm clearly out of my depth . . .

 

[NormPeterson]

@drwebb - the mid-engine Corvette is hardly an overnight happening. Rumors of near-future introduction of a mid-engine Corvette have been bandied about for at least 50 years. Closer to 60 if we're counting Duntov's 'CERV II' effort.


Norm

 

[NormPeterson]

BTW: Summing up point masses * distance^2 to get total inertia can yield a 40 - 50% error because you also need to add in the inertia(s) of the 'point' masses. ( Remember your parallel axis theorem ? ) For a dressed engine + transmission, this ain't something to ignore.

Understood.

But I think for these sorts of back-of-napkin efforts you can safely ignore the 'own inertias' of the masses that you're moving . . . as long as it's only the difference in MOI that you're looking for (in this discussion, those 1200-ish numbers). Of course, you'd have to include them if as part of the FE to ME changeover significant changes in the 'own inertias' of the things that were moved occurred.


Norm

 

[GregLocock]

Part of the reasoning may have been boredom. There are few things more expensive than a bored engineer. If they had to crank out yet another FR car they knew they'd be compelled to duct tape a few new bits and pieces onto that tired old structure. By going to MR they got to redesign lots of things and abandon some sacred cows, and have a completely new car to develop.

Cheers

Greg Locock


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

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

 

[CWB1]

OK, for the folks watching at home- so if there's no technical basis for switching the 'vette from front (front-mid) to mid (rear-mid)

Its the classic argument - its technically superior, but most owners are neither skillful enough nor in a position to use its potential so yes, its marketing driven as most luxury products are.

 

[gruntguru]

Exactly. What does a manufacturer do to promote their particular supercar? How many buyers track test before they buy? They all want to own the fastest thing around a track but have to rely on some scribe's opinion - or the lap time at the Nurburgring with a hotshot behind the wheel.

That - and the 0-60 (0-100k) time as I said earlier.

je suis charlie

 

[cibachrome]

It's not about satisfying buyers who go golfing with their image cars or grocery shopping or Dream Cruising down Woodward Ave.

When the company decides it wants to win at Daytona, Lemans, fastest lap at the 'Ring, whatever, in a World Market, and can predict the sales and price break, a platform is initiated. Cadillac is another GM division who wanted some racing DNA whether it sells or not on Monday. It strengthens the backbone of the whole division. When a person who has been on a racing team goes back to the bread and butter vehicles, the product only improves, faster timing, new tools, styles, quality, fitment, testing machines, analysis techniques, even the way they walk.

Management (Reuss) knows this.

 

[GregLocock]

I worked on a couple of start from scratch supercar programs, neither got even to prototype stage. They weren't especially interesting to work on once the crazier ideas got thrifted out, although I would have liked to have worked on the V12 with a central power take off, (essentially two V6s nose to nose). The crankshaft dynamics of that would have been interesting.

Cheers

Greg Locock


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

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