There are a lot of talented fabricators and car builders around here...I am not one of them. I could use some pointers on basic powertrain alignment relative to the chassis. I understand the importance of the pitch angle (rotation about the green axis) relative to the rear diff. My question is more about the "yaw" angle (or rotation about the blue axis.)
In mockup I've been trying to keep this at 0 deg or straight down the centerline of the car. However, it would fit much better in the engine bay and transmission tunnel if slightly rotated...lets say between 1-2 deg from parallel with the centerline. Is this a big no-no for driveline harmonics or other issues? Thanks in advance
1 or 2 degrees amongst friends?
How long is driveshaft?
That seems like something you could "correct" with a high-quality U-joint at the driveshaft.
In reply to Spearfishin :
1.5 deg measured in a rough CAD mockup, so who knows in reality? 
Short, rough guess is probably around 31"
I'm not a mechanical engineer, but I don't see why the driveshaft should know or care about the difference between misalignment in the horizontal plane and the vertical one.
In reply to codrus (Forum Supporter) :
That's my gut reaction as well but now I'm overthinking it and wondering if there are weird dynamic effects during rear axle jounce or something.
I'm probably just stalling on cutting half the car up 😬
codrus (Forum Supporter) said:
I'm not a mechanical engineer, but I don't see why the driveshaft should know or care about the difference between misalignment in the horizontal plane and the vertical one.
From my understanding it does, if you have some horizontal angle on the trans output you'll need some on the diff to match, due to driveshaft phase. This is why you never see production cars and rarely see race cars where the driveshaft isn't perfectly aligned with the longitudinal axis of the vehicle. Even if the drivetrain is offset, it will be aligned straight so that the same angle occurs at the diff.
That said if you use a CV driveshaft then you don't need to worry about angles anymore, other than staying within the limits of the CVs.
enginenerd said:
In reply to Spearfishin :
1.5 deg measured in a rough CAD mockup, so who knows in reality?
Many cars have the engines cockeyed in the chassis.
My most familiar example are first generation RX-7s. Rotaries are significantly wider on the right side than the left, but the chassis has the trans tunnel slightly biased to the right (I assume they designed the car for LHD markets?) so the engine is cockeyed by a few degrees so the exhaust can clear the RHD market cars' steering boxes. This doesn't seem to hurt much of anything. The fan shrouds look weird though.
It's also one more reason why a LHD RX-7 has a crazy amount of left front corner weight. Everything is on the left side of the car, even the engine itself 
Even in conventional driveshaft setups with live axles used for many decades when the rear axle moves up and down compared to the chassis many times more than 1.5 degrees. And thats with old fashioned u-joints. Just saying.
Does the engine have to be rotated, rather than simply offset to one side?
In reply to DeadSkunk (Warren) :
It would make my life a lot easier if I can have it slightly rotated. Offset but parallel to the chassis centerline doesn't help me much in this case
Pete. (l33t FS) said:
It's also one more reason why a LHD RX-7 has a crazy amount of left front corner weight. Everything is on the left side of the car, even the engine itself
Wow, I never paid much attention but that's interesting.
Sounds like the correct way to do this is to keep the engine/transmission centerline perpendicular to the axle to avoid severe driveline vibration. Just like with setting pinion angle, ideal would be to keep the angle of both U-joints equal. So when viewed from the top an offset of the engine/transmission centerline relative to the rear diff centerline is fine, so long as they are parallel. Sounds like what I was proposing is not ideal.
Have you ever looked under the hood of a Mercedes V8 powered Gwagen?
Use a double cardan or CV joint instead of a U-joint. U-joints are not constant velocity when operated on an angle. They speed up and slow down. The reason they work properly is because the rear U-joint is installed on the opposite (90 degrees out) plane and cancels the speed variations if operated on an equal and opposite angle. Watch this video:
This is why FWD cars have CV joints. They have to smoothly transfer torque in 3 dimensions; with suspension movement and steering. This is also why older 4wd trucks with a solid front axle can use U-joints, because they only have to compensate for steering angles.
You could use u-joints... but only if your pinion angle were offset equal and opposite to the engine's offset.
U-joints work best when they operate on a modest angle of around 2-5 degrees. Installing them at 0 degrees is a possible recipe for vibes since there is no active resistance. Operating them on a slight angle causes them to apply forces to the elliptical travel, much like placing your hand on a speaker reduces it's sound.
The trick is that the angles need to be equal and parallel. If the back of the transmission points down 2 degrees, the pinion has to point up 2 degrees while maintaining a 2-5 degree deflection at each U-joint. It doesn't matter what plane it's on; up/down or left/right. Rotational torque doesn't care about where you get those angles. In most applications, it's simple to adjust the transmission and pinion up or down to get what you want, but in your situation, you can't get that. The pinion will be straight forward with no adjustment possible, while the transmission output will be crooked.
I suppose if you used an IRS where the rear differential is separate from the rear suspension, you could install the differential crooked.
If I read this and did my finger pointing correct, you want to move the front of the engine to the left and the end of the tranny to the right 2°. Is this figuring out to be way too many inches on the end of the tranny? Is the engine not fitting in the bay straight?
enginenerd said:
Pete. (l33t FS) said:
It's also one more reason why a LHD RX-7 has a crazy amount of left front corner weight. Everything is on the left side of the car, even the engine itself
Wow, I never paid much attention but that's interesting.
I think you'll be fine as is. If you have issues, I like the idea of having a double Cardan joint at the transmission end but only if the rear U joint is at a zero angle, since a double Cardan is effectively a zero angle U joint as far as driveshaft rotational speed oscillation is concerned.
In reply to Curtis73 (Forum Supporter) :
Appreciate the well thought out and informative response. Will likely be a solid rear axle car so sounds like I will need to keep the powertrain parallel to centerline of the pinion when viewed from above and calculate the resultant angle of the pinion when viewed from the top and side views to make sure it's in that effective 2-5 deg range.
In reply to enginenerd :
You bet, friend. Like we mentioned, you could do a Cardan joint. Many GMs did this on their higher-end vehicles like Cadillacs. Not sure if you noticed in the video, when you have equal angles, the U-joints cancel each other out, but the driveshaft in the middle still goes a wee bit faster/slower. It's not enough to really make a big vibe, but cars like Caddys, Olds, and some Pontiacs added a Cardan joint (which is basically two U-joints in the same hub) to make it really smooth. If you scavenged one of those for your transmission output, you can get away with a crooked engine.
Here is a doctored image of how I like to set them up:
Here is a Cardan joint. It stacks two u-joints out of phase, so any speed up/slow down generated by the elliptical angle is canceled.
In reply to Curtis73 (Forum Supporter) :
RX-7s were cockeyed sideways about three or four degrees, though, and Mazda sold a few million of them. It looks like the damndest thing, the front of the crank centerline is about 2" off to the left and the trans end is a little to the right. Pinion is almost centered in the chassis.
now, when the trans mount fails and the trans end shifts further to the right, then it starts to get unhappy
In reply to Pete. (l33t FS) :
It sounds like it's a matter of ideal engineering to minimize nvh and address durability concerns vs. the "what can we get away with?" compromise for packaging. I think it's wild the RX-7 was designed that way and is maybe a case for this particular swap project still working out ok if the powertrain is skewed.
Appreciate the discussion...I am learning a lot here (as I always do)
