[brian_]

Re: 1974 BMW CSE

With no drive train our coupe weighs in at:

Front: 1020 lbs. (Drivers 576 lbs, Passenger 444 lbs.)
Rear: 1140 lbs. (Drivers 588 lbs, Passenger 552 lbs.)
Total: 2160 lbs.

Drive train approximate weight: 1010 lbs.

I was surprised that the total for the powertrain came out that high, but inline-sizes are heavy and this is old tech. The large bias to the driver's side is strange - the steering column and pedal cluster shouldn't make that much difference. The diagonal values suggest that the car is not straight, or the springs are unequally sagged... and neither is surprising at this age.

In rough numbers, our Tesla drive unit weighs 295 lbs, and the battery modules weigh 55 lbs each for a minimum of 550 lbs. This brings us to 855 lbs. Add to that the weight of our fabricated mounts, battery enclosures, battery management and charging systems, and it appears that we should be able to end up about the original weight of the coupe. If needed, we still have the option of lighting up other portions of the car such as seats, carbon-fiber hood and more. We are exploring some suspension options that could be lighter and more friendly as well.

Carrying only two-thirds of the Tesla battery does keep that part of the weight down, which is okay if you can live with the shorter range and lower motor power.

I doubt that all of those other bits can be done in 155 pounds, but it might be close. Remember to add a cooling system to your list. The structural changes to mount and drive unit and the body changes to accommodate it will also add weight.

It's very unlikely that you could save any meaningful amount of weight (relative to the total for the vehicle) by changing suspension components, at least at any reasonable cost.

 

[brian_]

Yes I was happily surprised at the weight. While my scales aren't completely accurate, they are accurate to themselves, and since my first weight was very close to what BMW says the car weights, I know that I am extremely close.

I'm not questioning the validity of the measurements - was just surprised by some of the weight information.

Regarding the LR difference, that inline did lean way into the passenger side, so it makes sense for the amount of difference there- though more than I expected. And with a full tank of gas removed from behind the rear passenger tire (and the spare tire still in) the rear balance made sense to me. I may ditch the spare, but wanted to leave it in for this purpose.

The engine inclination and fuel tank location would explain a right-side-heavy car, which isn't so bad because it counterbalances a solo driver's weight. They don't explain why the car with no engine or fuel tank and no driver is so left-side-heavy. The spare tire doesn't weigh much compared to the side-to-side bias.

The left-heavy distribution may not be a big deal, and the cause probably doesn't matter, but I think it's worth considering some placement of components toward the right side.

Regarding my 10 module minimum. It is enough to build up the voltage, but true, not nearly the range.

10 modules isn't just 10/16ths of a 16-module Tesla's capacity, it's also 10/16ths of the voltage. You can get Tesla Model S/X modules rewired to 12S (instead of 6S), but if you do that you can only use up to eight of them. Available voltage limits available power at higher speeds; that might not matter to you.

Still, the Model S is 4800LBS, and my car 3170. That is fairly significant to the performance and range I should expect. I do hope to get 12 modules in there, but I think my weight to performance ratio will top out about there. 16 modules in a 4800LB car that gives 300 mile range versus 10-12 modules for 3200 pounds and I should beat my 250 mile goal. I had the 70D with 240 miles and I was very happy for everything around So-Cal.

I understand the logic, but the expectations for range may be high. The Model S is much heavier than the CS, but consumption doesn't just depend on mass. The Model S is also larger, but likely has a much lower coefficient of aerodynamic drag than the CS (which was okay by the standards of the 1960's and early 1970's). Especially at higher speeds, the aero will probably be more important than the weight. Even if the range is only 200 miles, that would be long for a DIY conversion and presumably would be fine for this vehicle.

 

[brian_]

The diagonal values suggest that the car is not straight, or the springs are unequally sagged... and neither is surprising at this age.

A little more detail on this issue...

With a full tank of gas our CS weighs in at:

Front: 1743 lbs. (Drivers 916 lbs, Passenger 827 lbs.)
Rear: 1427 lbs. (Drivers 719 lbs, Passenger 708 lbs.)
Total: 3170 lbs.

The diagonal totals should be the same for a properly set up normal car (not a drag racer or an oval track racer). In this case:

• left front + right rear = 1624 lb
• right front + left rear = 1546 lb

That's only 78 pounds difference (2.5% of the total), probably no problem at all.

With no drive train our coupe weighs in at:

Front: 1020 lbs. (Drivers 576 lbs, Passenger 444 lbs.)
Rear: 1140 lbs. (Drivers 588 lbs, Passenger 552 lbs.)
Total: 2160 lbs.

In this case:

• left front + right rear = 1128 lb
• right front + left rear = 1032 lb

That's 96 pounds difference, and 4.4% of the total. The car appears to have been set up to handle the offset engine, and will likely need some tweaking after half a ton or more of EV components are added (in different positions from the original powertrain).

It's a relatively small detail, and I'm sure that most builders wouldn't worry about it, but since you have the individual corner weights and are building a car known for handling, it would be a good detail to sort out... eventually.

 

[brian_]

... I am excited to fit in the components in such a way to better the original weight distribution if possible. I will be trying under the rear seats as well as a bit out back- but using the engine compartment for a majority.

The car was moderately front-heavy (55%) when stock, partially due to the engine being too far forward; that wasn't bad for the time, is still normal for a non-high-performance car, and shifts rearward with passenger and cargo load (so you don't really want 50%/50% when complete and unloaded).

Without a powertrain, the roller is at only 47% front weight bias (because the boat anchor engine was up front), which is too rearward... but it will shift forward as those battery modules are stacked in the front, because that's best and almost only place to put most of them.

Under the rear seat would be nice, but since the fuel tank wasn't there (unlike a modern vehicle), and the suspension extends under the seat, there isn't a good place to receive big rigid boxes. I would be tempted to replace the rear seat with a substantial battery box under a parcel shelf, since I wouldn't expect to have rear-seat passengers.

The fuel tank and spare tire location is right at the rear, and not great for handling dynamics.

It's going to be an interesting packaging challenge.

There are some good drawings for layout planning online, such as those on this website:
BMW E9 CS group 2 (1970)

Another E9 project illustrates what you find when you strip the body down to just the shell, including a not-so-useful space under the rear seat (see the last photo):
BMW E9 – chassis restoration

 

[brian_]

A little more on the space in the rear...

The E9 seems to have the space under the trunk divided down the middle, with the spare tire well on the left and the fuel tank on the right. See the third-from-last image in the restoration link above for illustration. Whatever is left of that space after the Tesla drive unit could be used, but staying to the front of that space would certainly be preferable.

The spare tire well would need to be cut out and replaced with something better shaped for the battery modules. The fuel tank side is just a hole, so a battery box is required to protect the modules from the outside world, to protect the modules from stuff in the trunk on the inside, and to protect interior contents from the effects of severe battery failure. The interior side protection is needed on the spare side, too.

The structural divider between the spare and tank could probably go, but it's worth considering how to effectively replace it, for rear impact strength if nothing else.

The drive unit mass is substantial, and it is centred behind the axle, so you may not want any battery behind it all to avoid to front-heavy mass distribution. It's worth some specific planning and calculation to see what configuration will work.

There is another rear location which could be used for battery modules: the front of the trunk, on that shelf over the axle. It's high (which is bad), but further forward and a usable size and shape. If using that and not the fuel tank space, the fuel tank space could be made into a covered storage cubby to make the remaining trunk more usable.

 

[brian_]

Perhaps when the car comes back clean I might take a batch of photos at even angles and draw in to scale some of these ideas.

I'm looking forward to those.

 

[brian_]

Depending on your motor weight, you might get a better weight distribution by locating the spare wheel under the bonnet/hood, and using both halves of the under-trunk space for batteries. You'll definitely shift the CoG downwards.

An interesting idea. It would also shift the CofG rearward... maybe too much? I think it would make sense to calculate the new CofG - at least front-to-back and ideally vertically as well - before committing to component locations.

 

[brian_]

We started by clearing out the trunk floor...

It looks like a lot more than the trunk floor went.

In the middle, at the front of where the original final drive (differential) housing would have been, there appears to be a bracket which was likely bolted to the nose of final drive, and likely goes forward to a beam that the suspension arms mount to. In later photos, it looks like that bracket was cut off of the beam. The back of the final drive would have been supported by a mount. If that's how it was done, the beam and the final drive with its brackets formed a T-shaped subframe for the suspension, and the rear leg of that is now gone. With only two mounting points of the remainder of the subframe (the beam) to the car, the subframe is no longer properly supported and will twist under loads.

Ideally, the structural role of the bracket, final drive housing, and final drive rear mount, would be replaced by a tube arching over the drive unit to the new rear mounting point.

Maybe my guess about the structure is mistaken - maybe BMW decided to make one model different from all of the others - but do you have a photo of the suspension arms and what they are mounted to?

Until you're sure this is structurally right, I would encourage you not to support the car on the rear suspension, and certainly not to drive it.

The attached images:

1. image of cut out trunk floor, showing the bracket
2. later image, showing the subframe cross beam and where the bracket was probably cut off of it
3. suspension and subframe from a BMW 2002 (not a CSE), but is shows what the corresponding parts of the CSE likely looked like:
   (somebody else's image from Photobucket, plus annotations)

 

[brian_]

We have a large and deep rear section, then an upper forward shelf (with perhaps a plexiglass floor for any Cars & Coffee events we may find ourselves at).

I like that! Maybe place for a bolted-in panel, with steel (or aluminum) and acrylic (or polycarbonate) panels for driving and showing?

 

[brian_]

You are correct that we did cut off the bracket that held the front of the Diff, and the rear of the Diff did connect to the car forming the tee you mention. We do plan to reconnect this tee structure in some form.

Good

It is a bit of a strange setup, as the point of isolating this front unit (in my estimation) was purely to isolate the differential vibrations from the car. Front drive cars often simply attach the trailing arms to the car itself.

The T-shaped subframe has the same purpose as any other rubber-mounted subframe: to isolate the components on it from the vehicle structure. Yes, lots of vehicles do not use a subframe like this... they mount powertrain and suspension components directly to the structure at either at the front, or the rear, or both. Both differential gear noise/harshness and road noise/harshness through the suspension are handled by the rear subframe in this case; that's typical for rear-wheel-drive vehicles with an independent rear suspension, but not a universal practice.

So we are doing a bit of research to determine the best course modifying this. Either we reverse the tee shape forward and connect it with a mount (with bushing) in the tranny tunnel, or we simply shorten the tee and attach closer to the Tesla front mount.

Logically the Tesla drive unit would be mounted to the rear subframe; that's what Tesla does (both front and rear), and what probably every other current production EV has at the driven axle(s). Ideally, perhaps the existing CSE subframe could have been extended to two new rear mounting points to the unibody (replacing the differential case mount), and the Tesla drive unit could have been mounted to it instead of directly to the body. The Tesla front mount would be bracketed off of where the diff bracket was, the Tesla side mounts would go to the subframe ahead of the new subframe mounts to the body, and the rear Tesla mount would go to a new crossmember of the subframe. But that isn't the only way to go and it's not what you have been building.

Assuming that you want to stay with the direct-to-body mounts that have already been built for the Tesla drive unit, all that remains is to restore a third mounting point for the subframe, which now handles only the suspension. To be effective, it must be significantly away from the axis formed by a line between the mounts on the crossmember beam; the diff mount was well behind that (and a long and awkward S-bend tube could go all the way back to your new crossmember), but a replacement could be significantly ahead as you described, or even above. Maybe back and up to your Tesla unit front mount would work. The point is primarily to keep the subframe from rotating around that axis between the remaining mounts.

Thanks again for your thoughts and concern - it's what I love about this group!

Good. Some people really only want to share what they have done, rather than have a discussion of possible designs.

 

[brian_]

Don't forget that the Tesla unit will see wheel torque not motor torque

The further apart (fore/aft) that the mountings are the lower the loads that your mountings will see

That brings up another reason for subframes: to spread loads further apart on the body structure.

In some case there is no isolation at the points where the subframe bolts to the body (it's just solid), which indicates that the subframe is being used to carry loads to suitable points in the body structure... or in some cases to allow a large assembly of components to placed into the vehicle on the assembly line in one operation (such as an entire engine, transaxle, half shafts, hubs with brakes, and suspension into the front of many vehicles... or the drive complete drive unit and suspension into the rear of a Tesla).

In this case, the stock Tesla mounting points (where the drive unit mounts to the subframe) are being used, and they're at the extreme front, back, and left side of the drive unit. During acceleration the front mount will push up on the body mount bracket, which looks quite substantial. Presumably that beam added for the rear mount can handle a similar downward force. The left mount (assuming there's no right-side mount) will basically just see a small share of the weight of the drive unit vertically because the motor is on the left side (plus all mounts take lateral force in corners and longitudinal force when accelerating and braking due to the mass of the drive unit).

Assuming 601 N⋅m (443 lb⋅ft) of motor torque, and 9.73:1 gear ratio, the axle or wheel torque could be 5800 N⋅m (4300 lb⋅ft), so if the front and rear mounts are about 60 cm (2 ft) apart, the difference in vertical force at those mounts could be 10 kN (2000 lb... one ton).

 

[brian_]

We decided on this route as the simplest and most robust. We did certainly plan for the 2000lb. We aren't auto engineers, but we do a lot of structural work where we have to account for 7x the maximum expected load for rigging. You can probably tell by the type of steel beam we used in the rear

Yes, the structural work looks good. My comments were mostly for those other forum members who dismiss mounting these components are trivial, and don't understand the forces involved.

Regarding that subframe 3rd mount, I am doing a bit of research on just how much torque she will see. The 20" back or so that it attached in the past was because she experienced a lot of torque to spin the wheels. Now she does nothing but go along for the ride. The main thing is to lock her in so that the two main bolts aren't expected to do all the work- they would shear eventually. So I thought of going perhaps 12" or so forward in the tunnel and attach with a generic isolated motor mount, or even as you mention, straight up to our front Tesla mount (other than it being crowded up there).

While torque reaction from the final drive housing would be the major torque about the lateral axis, that's not that's going on. The subframe takes all of the horizontal load of the suspension; while the suspension arm pivots and the subframe mounts are nearly in the same plane, that alignment isn't perfect so any cornering, accelerating, and decelerating forces will tend to twist the subframe around those two mounts. The third leg doesn't need to be anywhere near as long as it was (to the rear of the final drive), but something is needed, and it sounds like you're headed in a viable direction.

 

[brian_]

Maybe I missed out , but , how much modules will you put in ? , all 16? to get proper voltage..

Just search this thread for "battery", and you'll get the evolving story of how many modules will be used. Yes, pack voltage is one concern.

 

[brian_]

Nice work on the subframe extension. I don't think it matters for this purpose how the images are rotated.

 

[brian_]

Re: 1974 BMW CSE

I can't get the embedded YouTube function to work- if anyone has some advice I'd love to hear it!

The forums embedded YouTube player - what you get with the YouTube icon in the graphic editor or the YOUTUBE commands - use Adobe Flash. That works for some users, but is blocked by some browser and operating system combinations, leaving only a big non-functional blank.

The straightforward links to YouTube videos work fine, so don't worry about (or try to use) the embedded player.

 

[brian_]

Re: 1974 BMW CSE

We explored various methods to mate the output of the Tesla to our BMW hubs. We couldn’t modify and reuse the original BMW axles as they would twist like paper straws under the high torque of the Tesla unit.

We turned to The Driveshaft Shop in North Carolina whose slogan is “We specialize in the impossible”, and for good reason. We sent them the inner portions of a pair of Tesla axles and a host of measurements. A few weeks later a beautiful new pair of Frankenstein axles showed up and bolted right in. The craftsmanship is stunning.

Beautifully crafted axles. They are completely new but the far right section which was out of a Tesla. The left end bolts to the BMW wheel hub and the right spindled end inserts into the Tesla drive unit.

Okay, the inboard stub is Tesla, but not the inboard CV? Do you know what CV joints were used? I wonder if they might be the very popular Porsche 930 type, a bolt-to-flange style for which a variety of flanges are available. What "level" in their axle system did you choose?

 

[brian_]

Re: 1974 BMW CSE

From my invoice:

"108mm Axle with 28 spline bar and 108mm CV's on both ends with Chromoly internals (Made with 14-1/4" 28/28 Spline Axle Bar)
108mm Head for welding to CV's and Driveshaft Trans yokes 1018 or 1020 steel
Machine / Weld Tesla Inner to 108 Flanges"

Thanks

Yes, it looks like the joints are the 108 mm diameter 6-bolt type as used widely in racing and modified applications. I think the 108 mm size is larger than what Porsche used for the 930 (1975-1989 911 Turbo, and the source of upgraded parts for many hot VWs), which is good because as impressive as that car was at the time, the Tesla drive unit can put out more torque. The 108 mm bits are Formula 1 worthy stuff (in the right grade, and with the high-speed boots).

The "bar" is the shaft between inner and outer joints, using the 28-count spline most common for these joints, regardless of what either the Tesla or the BMW would have had. The length of the bar is to suit the dimensions from the car and the width of the drive unit.

The Tesla stubs were machined off and welded to flanges, which in turn are bolted to the inner CVs.

The description doesn't explain what was done on the outboard end to mate the bolt pattern of the outer 108 mm CV body (six M10 bolts on a 94 mm diameter circle) to the BMW's stub axle. The CSE appears to use this style of CV joint; it seems unlikely to me that it would be the same size, but presumably this axle specialist has the sizing right.

 

[brian_]

Re: 1974 BMW CSE

Yes, that I all understood, it was your question of the CV that I really didn't know. The 94/108mm is the stock BMW hub. They used the 108 at the inner as well to match them up.

The breakdown was as much an exercise in understanding the setup for me and getting confirmation, as information for anyone else

After posting my comment I found the GKN Löbro website with their catalog, which has an excellent description of the various axle configurations which their parts support and which they can supply. In their illustrated list of axle assembly types, this is a Type 102, the first one shown... plus that custom Tesla adapter as a (in GKN's terms) "connecting device".

It's fortunate that the BMW uses such a common CV joint type and size, making this straightforward solution possible.

I'm more accustomed to front wheel drive vehicles, which have a non-plunging high-angle CV joint on the outer end and a tripod joint for high plunge on the inboard end, and rarely have a bolted flange anywhere in the shaft assembly; rear wheel drive like this tends to use a plunging low-angle joint at each end, and - I now realize - often have four bolted flange connections in each shaft.

Clearly the CSL racing variant of the coupe that dominated in the early 70's was a beast. I assume they shared the same hubs as the street cars like mine, fortunately for me. I actually asked Lee about that while I was sending him all my measurements: "Will the hubs be a weak link" and the answer was no.

That is a nice feature. It turns out that according to the GKN Löbro catalog the 108 mm joint is "size 15", only midway up GKN's range of sizes and capacities.

Thanks again Brian, you sure know your driveline That's what I learned about this group early... I assumed it would be full of "EV" folks more than car historians and enthusiasts, but I have been proven wrong many times. There's a great diversity of skills here across all aspects of autos, old and new.

There certainly is diversity, and I agree that's a good thing!

 

[brian_]

Wow, Greg the brake lever is tall! It would be nice to be able to move the brake pedal pivot to inboard of the steering column, with the pedal shifted inboard as well (but less) to follow the accelerator pedal. If Greg were shorter, he would be lighter and stiffer... and it would be nice to have less gap between the two pedals which are both run by the driver's right foot. But even at this length, it's an interesting solution.

The more common way to transfer pedal action, used commonly for clutch actuation (before everyone went to hydraulics or cables) and even to run a brake master cylinder on the opposite side of the car from the driver (due to conversions between left-hand-drive and right-hand-drive), is a horizontal shaft with a crank on each end (the one on the pedal end can be the pedal arm itself). Instead of bending stress on a long beam, it has torque on a tube; instead of twice the pedal force applied to one pivot, it has the pedal force applied to each of two bearing locations. More common doesn't necessarily mean better, of course.

 

[brian_]

A little bit of logo work and the box looks very nice in the car. She comes in around 750 lbs. That’s about what the stock drive train and accessories, so the car is comfortable with the weight.

The front battery pack is similar to the engine and transmission, but the drive unit and rear battery pack are more than the original final drive and fuel tank, so I assume that there is a significant but manageable increase in total weight.

The balance on the turns felt nice, but we were keeping it slow so we don't know nearly enough there yet. We know we have more weight beyond the axles than we would prefer, but our goal of a 200 mile range car required some compromise. Very interested to see how she does at speed in corners.

In mass distribution, that shift outward is the handling concern: the centre of the front battery pack is further forward than the centre of the mass of the engine and transmission, and the drive unit is further rearward than the final drive and probably even further rearward than the centre of the final drive plus fuel.

This is a pretty common consequence of a conversion. As long as expectations for handling are moderate, it should work fine.