[gunns]

I already have an email out to a CV axle mfg but I haven't heard back yet; work has gotten in the way of me sending out some more. In the meantime, I wanted to see if you folks might have some ideas on what's the maximum angle I can expect out of some CV axles and still have them live? I have experience tearing axle boots and destroying axles at max drop when offroad racing so I know there's such a thing as "too much" for a given application. This car won't be going over whoops BUT I do want to confirm that this is viable before cutting any subframe bits.

Here's the math I've done so far.

• The distance between the LDU and the spindle is 20"
• At max drop (coilover suspension unloaded so the equivalent of the Telsa mid jump), the offset in the horizontal plane is 6.5"
• At max drop, the offset in the vertical plan is 3"
• By my math the max offset is the angle between 20" and the hypotenuse of 6.5" and 3" (7.16").

This means the max angle is 19.695 degrees

I can move the motor forward an 1-2" (horizontal plane) by cutting off some frame mounting points which need to be relocated anyway.
That would reduce the angle to 16.254 degrees

Q: Is this too much?
My internet searching says that <20 degrees should be good enough but I just wanted to double check.
FWIW, the EV motor makes 435HP and the equivalent amount of torque.

 

[remy_martian]

What's that frame from and what body mounts to it?

 

[gunns]

83 Ferrari Mondial. Midengined car originally.

• At the forward part of the subframe, there are 16 M12 bolts on 4 flanges holding the rear subframe to the frame under the body/cabin of the car.
• There are 4 more bolts holding the subframe at the rear to some metal supports that look more like rear end crash structures than load carrying surfaces (2 connect to the same bracketry that the mufflers originally hung from.
• Im going to have to lift the floor of the rear trunk by ~3" to allow for clearance of the LDU. This necessitates relocating the two body-side mount points that the two bolts at the top anyway. You can sere them on the pictures at the two tabs with holes that would hold vertical bolts.

The alternative is to cut two of the key structural supports that hold the shock/coilover top to the subframe and trusting myself to reinforce the support sufficiently. Otherwise, short of using something silly/overengineered like portal axles, there is no way to line up the LDU axle take-off with the spindles. I'd rather move the trunk upwards and cut a non-loadbearing bracket.

Anyway, do you have experience with maximum CV axle angles?

This is what I found so far but I'm curious if someone has direct experience
from:

CV joint tech -- A primer on constant velocity drive joints and diagnostic tips

I realize that much of this introductory text will seem very basic, but let’s start by explaining CV joint technology. A CV (constant velocity) joint is a torque/drive mechanical coupling in which the rotational speed of the output shaft (inner joint) matches that of the input shaft (outer...

www.autoserviceprofessional.com

Outboard joints

Outboard (wheel side) CV joints are usually the “fixed” type, also often referred to as a Rzeppa joint. This features a bearing race with slightly arced and offset longitudinal grooves. A series of ball bearings run along these grooves. The balls are “trapped” in a cage that keeps the balls aligned to their grooves. The offset groove/caged ball design allows the joint to articulate to follow suspension travel and steering angles. These are also referred to as AC (angular contact) joints. Depending on the specific manufacturer, a typical outboard CV joint will allow a maximum articulation angle of about 47 to 50 degrees (naturally, a higher maximum angle will allow a smaller turning circle).

NOTE: Because of the greater operating angles and their greater exposure to road hazards (rocks, dirt, moisture, etc.), outboard CV joints tend to wear sooner than inboard CV joints.

Inboard CV joints

In order to accommodate suspension travel, the inboard joint is designed to allow in/out or “plunging” movement in addition to articulating for suspension travel angles. This plunging movement allows the shaft assembly to change its length during suspension travel, compensating for control arm up-down/angles and to prevent restricting lower control arm movement.

Several styles of inboard CV joints are in use (although they all accomplish the same task). Tripod style CV joints use a three-legged design of three equally spaced roller bearings (instead of balls) that glide along track grooves inside a “tulip” style housing. A DO (double offset) is also a plunging type joint, but features a series of ball bearings.

Typical inboard CV joints provide a plunge movement of about 50mm and a maximum articulation angle of about 22 to 31 degrees (depending on make and model).

 

[Electric Land Cruiser]

Is that a 3D printed mock-up of a Tesla drive unit? Super cool!

 

[gunns]

Is that a 3D printed mock-up of a Tesla drive unit? Super cool!

Yes. Took 30 pieces because of printer limitations. It helps to have a friend who went big into 3d printers and became the "dominant" supplier of coronavirus christmas ornaments. I'm not sure it would be cost effective to produce for resale because its a good amount of material and print time unless it was priced at MORE than what a fake V8 mockup block typically sells for (FAKE P-Ayr Chevy Small Block LS1 Mock-Up Engine Block - FREE SHIPPING)

I can say that moving this 15-20ish LB block of plastic is a ton easier than hoisting up a real LDU.

---
BTW, I finally heard back from driveshaftshop.com:
If using 930 CV’s which would more than likely be the case for this they can run up to 25degs before the Boots themselves start to fail.

Another friend who is an ASE mechanic and does lots of custom work advised that FWD axles still work at opposite lock and they are at a MUCH more extreme angle that what I'm worried about. His concern for longevity is to get it as straight as possible when the vehicle weight/batteries/etc are on it. I won't know that until I reassemble things and with coilovers, I suppose I can also adjust ride height as well. If they aren't binding at max drop, you'll be good

--
Next stop is to start cutting.

 

[remy_martian]

You usually start off with zero angles and deviate during suspension motion only. Otherwise the CV joint rollers are traveling a huge distance per rev. The Model X has had problems with front CV joint angles from its initial launch, so getting it right stumps even Tesla's seasoned suspension engineers. You may have to repack the grease every 5,000-10,000 miles, for example.

If it was my project, and I didn't want to rework the suspension mounts, I'd relocate the Tesla inverter, push the DU into axle center, and get the joints zeroed out, looking from the top.

The 1962 Corvette thread talks about doing this, and it has been done on a C5 Corvette build by someone on the forum (see the '62 vette thread again) already on a completed, driving, car. You'll be happier with the handling with the tighter polar moment.

Another option, one I would do before cracking the inverter, but which it sounds like may be beyond your skill level or willingness, would be to use the Mondial's pulled subframe to make a jig to locate key features and fab a new subframe in the jig with the Tesla LDU fitted.

Having the CV joints all spazzed might work, but that's not what they're for.

 

[gunns]

You usually start off with zero angles and deviate during suspension motion only. Otherwise the CV joint rollers are traveling a huge distance per rev. The Model X has had problems with front CV joint angles from its initial launch, so getting it right stumps even Tesla's seasoned suspension engineers. You may have to repack the grease every 5,000-10,000 miles, for example.

If it was my project, and I didn't want to rework the suspension mounts, I'd relocate the Tesla inverter, push the DU into axle center, and get the joints zeroed out, looking from the top.

The 1962 Corvette thread talks about doing this, and it has been done on a C5 Corvette build by someone on the forum (see the '62 vette thread again) already on a completed, driving, car. You'll be happier with the handling with the tighter polar moment.

Another option, one I would do before cracking the inverter, but which it sounds like may be beyond your skill level or willingness, would be to use the Mondial's pulled subframe to make a jig to locate key features and fab a new subframe in the jig with the Tesla LDU fitted.

Having the CV joints all spazzed might work, but that's not what they're for.

• Pushing the DU into axle center on the horizontal plane requires reworking the suspension mounts. I'll get a decent amount closer by cutting the two top subframe mounts but i wanted to sanity check to see if AS-IS this was workable.
• I plan to go through the vette thread in far more detail.

My goal is to get this car running as an EV first. If I don't like the polar F/R weight distribution, I can always redo things.
I'm also going to be pretty realistic about this car's use case: this car isn't a track weapon or even a car I plan to autocross. It's most practical use-case is to get groceries, drop off/pickup my kid from school, and maybe drive around town on the weekend. I'm not going on roadtrips with it.

As far as building a jig for the subframe, this is a pretty silly suggestion. If I'm unwilling to cut the existing subframe load bearing support and engineer/bubba-in a replacement support, I'm unwilling to go through the effort (and unwilling to pay someone) to fabricate my own replacement subframe as a first step when other mounting methods are possible AND this car isn't an EV yet. Much like it doesn't make sense to harvest your own wheat if you first want to learn to bake bread, Id rather have the mondial running first and then decide if its adequate or not. If the first loaf sucks, I'll start moving up the development chain (milling my own flour, harvesting my own wheat, whatever).

Good news is this is NOT a project for someone else w/ a deadline OR a vehicle I need to get running for my day job. Its a hobby.

 

[remy_martian]

Relocating the inverter does not require any suspension rework at all...the diff of the DU winds up off center longitudinally and you have unequal length axles.

The Tesla DU is an ugly conversion choice because of its width - unless you lop the inverter off.

 

[gunns]

Relocating the inverter does not require any suspension rework at all...the diff of the DU winds up off center and you have unequal length axles.

Holdup - maybe I'm not understanding your suggestion option.
I know there's an inverter board inside this LDU. What are you proposing I relocate exactly?
How would that get me extra distance between the "barrels" of the motor windings and the axle takeoff point on the gear reduction unit?
If I flip it around, I have a physical issue where the barrels of the motor winding will hit the forward support of my shock top mounts.

This pic depicts the problem. I could scoot the LDU back another inch or two but you can see where the axle takeoff point (right where the two colors of plastic meet) runs into the key suspension point.

Q: What are you proposing exactly?

 

[brian_]

I would have suggested GKN (which has a good full-line catalog available online), but you already found the relevant information from another source. CV joints can withstand a lot of angle, but for long-term durability you want to use that capability only which turning (not relevant here since this is a non-steering axle) and with extremes in suspension movement, not continuously.

Having said that, some vehicles are designed with significant (a few degrees) static CV joint angles, for packaging reasons similar to the issues discussed here. An old example that comes to mind is that when VW jammed an inline four-cylinder into the T3 Transporter (Vanagon) that was designed for a flat-four, they did it by shifting the transaxle forward and forcing CV joint angles that I found surprising, but apparently worked. Some vehicles with raised suspension also run significant static angles.

 

[gunns]

I would have suggested GKN (which has a good full-line catalog available online), but you already found the relevant information from another source. CV joints can withstand a lot of angle, but for long-term durability you want to use that capability only which turning (not relevant here since this is a non-steering axle) and with extremes in suspension movement, not continuously.

Having said that, some vehicles are designed with significant (a few degrees) static CV joint angles, for packaging reasons similar to the issues discussed here. An old example that comes to mind is that when VW jammed an inline four-cylinder into the T3 Transporter (Vanagon) that was designed for a flat-four, they did it by shifting the transaxle forward and forcing CV joint angles that I found surprising, but apparently worked. Some vehicles with raised suspension also run significant static angles.

One problem is that I'm not used to buying custom axles so I had to rely on an axle vendor to tell me what sub-components make sense. I'm curious about how I go about ordering these things.

On one side, I have the Tesla gear reduction unit. Some companies sell stub axles that terminate in the 930 CV

Tesla 930 Stub Axle Conversion Kit for Porsche, VW and Others, EV West - Electric Vehicle Parts, Components, EVSE Charging Stations, Electric Car Conversion Kits

On the other, I have the 83 Ferrari Mondial spindle. It seems to have a stub axle that terminates in this flange.
Q: Is this a standard flange?

Since I have to get a custom length axle anyway, will a custom axle shop sell me a pair of axles in the right length with the mondial compatible flange on one end and the tesla splines (google says 29 splines) on the other?

 

[remy_martian]

Take the inverter barrel off and remote it.

 

[gunns]

Take the inverter barrel off and remote it.

Your comment made me start digging into the LDU into more detail.
I see now that one of the barrels holds the windings and the other holds the inverter.

Advantage
- I could then push the motor forward to directly line up with the spindles in the horizontal plane.

Disadvantage

• Added complexity. What you are proposing requires, as far as I can see, the fabrication of 2 waterproof plates (one for the gear reducing unit and one for the new inverter) along with 3 HV cables to connect the two pieces. Additional brackets would need to be fabbed but again, entirely possible.
• I also would now be dealing with unequal length axles (quick measurement shows something like 26" and 14" vs 20" / 20").

• Seems like that significant delta would require thinking about maybe different axle thicknesses to address differences in torque.
• Would also need to think about how this would affect the Quaife LSD.

Your proposal does merit consideration and is within my fabrication capabilities. Does not seem to be a slam-dunk solution though because it adds both complexity and potential disadvantages. It's something to ponder.

P.S. You are a big fan of Hemmingway, aren't you?

 

[remy_martian]

Bingo!

There are several of us interested in tooling those plates to remote the inverter can, and most of the heavy lifting is being done by an actual ME, @428RC ...again, that lives in his '62 vette thread.

 

[brian_]

One problem is that I'm not used to buying custom axles so I had to rely on an axle vendor to tell me what sub-components make sense. I'm curious about how I go about ordering these things.

On one side, I have the Tesla gear reduction unit. Some companies sell stub axles that terminate in the 930 CV

Tesla 930 Stub Axle Conversion Kit for Porsche, VW and Others, EV West - Electric Vehicle Parts, Components, EVSE Charging Stations, Electric Car Conversion Kits

On the other, I have the 83 Ferrari Mondial spindle. It seems to have a stub axle that terminates in this flange.
Q: Is this a standard flange?

That looks like a typical CV joint flange for joints of the style which includes the "930" size, which is not surprising for this car. They're really properly identified by the outer diameter of the joint housing, but the 108 mm size is commonly called "930", I assume because it was first used in the Porsche 930 (a.k.a. Porsche 911 Turbo). You can check dimensions (bolt circle diameter, bolt hole diameter, and outer diameter) for compatibility, if you can find specs with the bolt circle and bolt diameter information.
Common VW CV Joint Specs
930E Extreme Angle CV.JPG - drawing from specialty joint supplier suggests that 108 mm housing ("930" type) common joint housing uses 94 mm bolt circle diameter and 10 mm bolt holes

A comment in a post in another forum (for a kit car using these joints) about "bolt circle diameter is 94mm and a 10mm bolt is used (108mm OD CV joint)" confirms the information from that extreme angle CV joint drawing.

There are various widths (axial lengths) of 108 mm joints of generally the same design, and they have different torque capacities, maximum angles, and plunge distances. With the Tesla tripod-type inner joints, the outer joints don't need to plunge at all... ideally they would have zero plunge; when using the "930" and similar style joints at both ends (rather than tripod inners), it appears common to use plunging joints on both ends (which seems like a bad idea to me).

Since I have to get a custom length axle anyway, will a custom axle shop sell me a pair of axles in the right length with the mondial compatible flange on one end and the tesla splines (google says 29 splines) on the other?

That's sort of a three-part problem: an outer joint compatible with the Mondial flange, an inner joint with stub axle that fits the Tesla splines, and a custom splined axle shaft bar between the joints. As long as the diameters of the two splined ends are reasonably close and both splines are common enough, then yes some shops can produce that custom bar. With flanges for 930-style joints on both ends, there certainly are suppliers which can provide a complete assembly of a custom length.

 

[brian_]

• I also would now be dealing with unequal length axles (quick measurement shows something like 26" and 14" vs 20" / 20").
• Seems like that significant delta would require thinking about maybe different axle thicknesses to address differences in torque.
• Would also need to think about how this would affect the Quaife LSD.

The usual design is a thicker or even hollow (and much larger outside diameter) shaft for the longer side, both to match the torsional stiffness of two shafts of different lengths, and to provide bending stiffness to the longer shaft to keep it from whipping. At the degree of refinement of a project like this, I'm not sure if this is necessary.

The other approach to offset final drives is an extended stub axle on the motor side, supported by a steady bearing, with a jointed shaft from there to the hub which is more similar (or actually identical) to the one on the other side... this is the Nissan Leaf design.

I don't think that it would matter to the differential.

 

[EV_Jibareaux]

I'm not sure it would be cost effective to produce for resale because its a good amount of material and print time

well you've got one, make a mold of that one and blowmold with HDPE, do on demand manufacturing to skip inventory taxes

 

[remy_martian]

Actually, vacuum forming them would be easy.