An engine produces 500 lb-ft. of torque at the flywheel.
The transmission is in 1st gear, which is a handy 3.0:1 ratio.
The ring and pinion ratio is an equally handy 4.0:1.

What's the load on the nose/pinion flange of the differential?
What's the load on the end of a 4-foot torque arm attached to the differential?

A Cobra Mustang has no torque arm (with IRS, you don't need one) and instead mounts the nose of the differential with an ear/tab on either side bolted to a brace. Why doesn't it just snap those tabs right off if the load is so extreme without a torque arm to spread the load?

An F-body has a torque arm which transfers the load to a mount at or near the tail of the transmission. Isn't a 4-foot torque arm like giving the differential a 4-foot long cheater bar to pry with? Or does it reduce the load by moving the leverage point farther forward?

A hypothetical conversion has an IRS differential and a torque arm to support the nose of it, mounted to a brace used to mount the transmission. Eventually, metal fatigue causes the failure of the mounting points, which were only intended for a lateral chassis stiffening brace. Is this simply because of the torque arm working up and down on acceleration/deceleration, or is it because load at the front of the torque arm is magnifying the load at the differential?

Just curious.

 

alright jim;
as a man with a torque arm in the driveway.......i want to see this turn into an advanced tech thread thats interesting.
feel free to use my car hypothetically...600hp,guesstimate 650 tq, [3.06 first gear, cant remember what 2nd is on a 700r] i have a spohn arm that is mounted about 8 inches rearward from stock on it own xmember.i cut a 1.6 60 foot with 3.73 gears [optimum would be 410 as i shift at 6400] but the 3.27s i have now im at 2.0-2.2 60' times,although my trans was wasted and i dont think i was heating the tires properly as a result...
as far as your theoretics, they intrigue me but i will have to refrain from answering until my state of mind is returned ...it is my sincere hope that this turns into a useful discussion with many outlooks that is the reason i float around this board.
jimlab--if you would like more specs and measurements to acurrize this theoretical ?, i would be happy to give you mine[and hopefully get a better combo resulting]

stealth

 

given that the axis of rotation would be the centerline of the axles under acceleration (pinion tries to rise). the tq arm reduces the load placed on the attachment point at the car. if you ignore the lca's of a 4rth gen and assumed the tq arm ate up the full load, it should be pretty easy to calculate the load at the attachment point based on thrust at the tire contact surface and working backwards. but due to deflection and exactly when in the event you're calculating real world results will vary. i'll leave the rest to somebody else i'm exhausted right now (frickin morning classes).

 

It reduces the load. The stock torque arm front mount would be broken pretty quickly if the torque arm was much shorter.

For a quick & dirty number the vertical load on the front of the arm is just the axle torque divided by the length of the arm--resulting in 1500 lbs for your example. If one wants to get more complicated, the height of the rear LCA mount will make a difference, etc, but since this can be changed significantly very quickly, ignoring it is usually a good idea when designing an arm--giving you a conservative result. You need a pretty big margin of safety anyway with the shock/inertial/etc loads that are difficult to predict. The toughest thing about a torque arm is that it's a beam in bending, unlike the other links. The load at the front mount is not usually where your biggest worry is--it's the moment right at and in front of the axle.

As for the Mustang IRS, it's really the red-headded stepchild of IRS designs so I wouldn't look to it for much inspiration. The front brace you mentioned can and does break--there are numerous aftermarket replacements and Ford has redesigned the piece. Also keep in mind, it isn't reacting all the torque by itself as the diff attaches to the cradle in other locations.

The hypothetical version you mention isn't going to accomplish much. In order for an IRS to operate properly, the diff needs to be securely mounted to the frame--using a 4' lever on the front as a part of that isn't going to be the most efficient way of doing it and since it won't work by itself it will be nullified anyway. In other words, in the end, the differential torque will simply be reacted in the chassis as a bending moment like any other IRS and won't affect the load on the tires--it's not really a part of the suspension geometry kinematically as it is with a solid axle (where the torque is reaced by the suspension links). It's best just to bolt the diff to the frame in the most rigid manor possible.

 

The load on the end is determined by how hard ya hook it and what the pinion angle is.

I have seen a TA about a foot shorter than a Spohn(home made) and it worked OK on a low 10's car..It's not so much the length(not real short like 18") as the angle of the TA in relation to the LCA's "ie" instant center which is in relation to the CG.

 

Free Body Diagrams are your friend.

 

Torque arm load at the end will be in proportion to geometry and torque at the wheels. Peak torque has to be considered if your're designing a connection. I would just guess that peak torque on a stick car at launch can be as much as 4 times higher than a calculated torque taken from a dyno curve and appropriate gear ratio's. Automatic cars have additional torque multiplication from the torque converter, but won't typically have as much instantaneous peak torque as the rotating inertia of a heavy flywheel and clutch suddenly engaging. Automatics with a transbrake probably fall some where in between. As said in an earlier post: A good free body diagram is your friend. FWIW the stock F-body Torque arm is not designed to take 5000 RPM M6 launches on slicks. Don't ask me how I know

Steve

 

Yes, assuming you "speak their language".

Doesn't the C4 (84-96) Corvette use a torque arm on the IRS?

 

c4s use two "support struts" for lack of a better term that go from the rearend to the frame.they arent quite "torque arms" in the fbody sense but theyre not really like control arms either..?..
more like links to control fore/aft movement and locate the rear.like half a 4 link mounted the wrong way.

 

It has plenty of faults, the most significant of which is that they compromised the rear suspension and differential by using a double bushings system to increase ride comfort. The differential and control arms are mounted to the subframe with rubber bushings at front and rear, and the subframe itself is mounted to the car with more bushings. It's no wonder they have terrible wheel hop issues and a lot of drivetrain noise.

However, all I was really talking about was the front mounting tabs/ears, which seem more than strong enough to put up with the load when rigidly mounted in the Factory Five Cobra chassis and in my hybrid setup, which borrows from that design...

http://home.gci.net/~jimlab/images/Cobra/PC080019.jpg
http://home.gci.net/~jimlab/images/Cobra/PC080024.jpg

The flimsy brace itself, yes. The tabs on the snout of the differential are what I'm talking about. One would think that if the load were as great as maximum torque at the axles that they'd snap right off. It is, after all, only cast aluminum.

http://home.gci.net/~jimlab/images/Cobra/P9010008.jpg

Precisely. The argument was whether a rigidly mounted IRS differential benefited in any way from a torque arm and what sort of load it was putting on the nose of the torque arm that would cause it to shear off body mounts originally intended for a lateral cross-brace.

The torque arm proponent (who makes them for no better reason than the F-body had one and when dropping an LS1 and T56 into an RX-7, you, um, might as well have one) says that the torque arm "stores" energy which it then releases over time, helping to propel the car. Putting aside for the moment the fact that he's barely competent enough to weld two pieces of metal together with good penetration (several of his torque arms have failed and had to be "recalled"), you realize quickly he doesn't know what he's talking about.

My argument is that the "sawing" motion of acceleration/deceleration with a rear-mounted differential that can pivot on the mount bushings loads and unloads the nose of the torque arm sufficiently to eventually cause metal fatigue on the flimsy transmission tunnel mounts and failure (which we've now seen on a few cars).

So we've got the torque arm conversions, the Cobra IRS conversions, and now two others (one a local to this board) have created completely rigid mounts (no bushings at all) for the front of the stock RX-7 differential that mount to the same type of mounting points as the torque arm cars (lateral cross brace), but only ~1 foot in front of the differential. It's my belief that despite the fact that it's apparently held up for a trip to the track and some 10-second runs that eventually it will suffer the same failure problems.

My setup, on the other hand, directs all of the differential load into the rear cradle and subframe, the strongest component on the underside and rear of the car. I just wondered what the maximum load is on the snout of the differential. I suspect that if you were using slicks that it might momentarily spike to whatever maximum torque at the axles was and then drop, but on a street car, you're disipating a great deal of that energy by transfering it into forward motion (or wheel spin), and you'll likely never see peak load at the front of the differential.

Thanks, that's what I figured.

 

The links that you're talking about (see picture below) are part of a pseudo-4 link setup for locating the axles and wheel, but they're not part of the differential mounting system. The axles themselves are a structural member of the suspension, acting like a lower control arm. The Jag IRS is very similar in construction.

The C4 uses a "bat wing" aluminum brace to mount the rear of the differential (with a bushing at the end of both wings), but it does have a short cast aluminum torque arm that rigidly aligns the front of the differential to a mounting point a couple feet forward.

I was unable to find a picture of the torque arm with a quick search, but I have several at home. The picture below shows the bat wing at the rear (mounted to brackets on the outside of the frame rails), and they've replaced the front-mounted "torque arm" with a bracket that mounts directly to the inside of the frame rails.

http://rodandcustommagazine.com/tech...mage_large.jpg

 

i see, said the blind man, to the deaf dog.

 

Sorry Jim, I was having intarweb problems.
I can help you with that. The diff is mounted in front and in the back. You need to measure the distance between these two mounts. Now, looking at the diff housing from the left side, idealize it like so:

http://www.jonaadland.com/NewPics5/DiffHousing1.JPG

That will give you the vertical loads at both the front and rear mounts. It's slightly conservative--assuming the mounts can't react moment, but generally that's the way to go. Start counting on those things reacting moment and it's a slippery slope which usually results in a flexi-flier, fatigue-cracking design....

So, it's a fairly large load. Before worrying about the aluminum tabs, let's focus on the subframe. The Mustang IRS isn't just held from the bottom, but also the top:

http://home.comcast.net/~ccjcobra99/...mmirsbrace.jpg

As you can see, it's pretty damn beefy. However, even after replacing the bushings with something more solid, there is still quite a bit of flex. Here's a simple mod William Mathis (long time Mustang suspension guru, famous for his books but has done a bunch of good work at Steeda the last couple of years) recommends to reduce it:

http://www.jonaadland.com/NewPics5/irs3.jpg

Basically further stiffening of the subframe connection to the unibody....

That brings me to my concern over the front brace of your subframe (which is very nicely done BTW). If it attaches to the body of the car somewhere in the middle that can't be seen from the pics you posted, nevermind the below. If not, I'm worried you're going to have quite a bit of deflection there. Looking at the front brace from the front:

http://www.jonaadland.com/NewPics5/FrontBrace1.JPG

As you can see, the deflection is a function of the cube of the length of the brace. Applying a perpendicular force to the middle of a long tube is a difficult thing to do well without adding a bunch of weight.

Which brings us to why the aluminum tabs on the diff itself are probably the least of your worries. It isn't the vertical forces themselves in pure shear that are tough to deal with--it's the moments resulting from applying those loads a distance away from where you're reacting them that causes problems.

Eccentricities are a bitch.

The aluminum tabs deal with large vertical loads but there's only a tiny bit of distance to create any moment on them. Mostly, pure shear. Like the "tip" of a torque arm--it doesn't need to be so beefy because there isn't a huge distance turning that load into a huge bending moment.

If you want to run some numbers on the brace given the above equations, "I" for a round tube is pi*r^4/4 for the outside radius minus the same for the inside radius. "E" is around 29 million for most steels off the top of my head.

As for all these "torque arms" etc, people are hawking...sounds like most aftermarket parts manufacturers to me--they don't know what the hell they're talking about. As you can see from the first pic, the net vertical force on the chassis will be zero. Extend the front mount out a couple of feet with a "torque arm" and it's still zero. Unless the diff can move up and down with the wheels, it simply isn't part of the suspension. It's part of the drivetrain--and angles, deflections, etc, in the drivetrain are important in their own way...but it's not a part of the suspension like a real torque arm on a solid axle is. It's never going to net a vertical force on the chassis unless it's free to move up and down with the wheels. The same thing drawn above on a solid axle results in the front reaction being provided by the weight of the chassis and the rear reaction being provided by the ground. There's a net force between them that lifts the chassis relative to the ground. With an IRS diff, both reactions are supplied by the chassis and the ground doesn't even know it's happening anti-squat-wise.

 

Thanks Jon. So let's say the cross bar is 30 inches long and the engine is producing 600 lb-ft. of torque with a 2.97:1 first gear and 3.55:1 ring and pinion...

I based my configuration on the Factory Five Cobra chassis, which now that I look at it again, has a shorter crossbar under the differential (the Mustang Cobra mounts the differential from the top only in stock configuration, the lower brace shown in your picture above is aftermarket) which you can see here. It's also boxed, not tubing, and has no bends in it. Space was at much more of a premium in my car.

http://home.gci.net/~jimlab/images/C...iveshaft-1.jpg
http://home.gci.net/~jimlab/images/Cobra/picture8.jpg

And here are a few more pictures of my setup...

http://home.gci.net/~jimlab/images/Cobra/P8210035.jpg
http://home.gci.net/~jimlab/images/Cobra/P8210040.jpg
http://home.gci.net/~jimlab/images/Cobra/P8210054.jpg
http://home.gci.net/~jimlab/images/Cobra/P8210047.jpg
http://home.gci.net/~jimlab/images/Cobra/PB120004.jpg

The revised version doesn't have the exhaust clearance notch shown on the left in the next to last picture. One of the goals was that the cradle be completely bolt-in so that you could revert back to stock if necessary.

Since my cradle was built, someone else did this...

http://home.gci.net/~jimlab/images/Cobra/DSC01419.JPG
http://home.gci.net/~jimlab/images/Cobra/DSC01427.JPG