Brand New Incredibly Strong One-Piece Transmission Input Shaft!!

[Nicholas R]

One of the first questions I always get when people see my car with the engine swap is "what about the transmission"?

The truth is, ever since I first started thinking about doing an LS1 swap, I've had concerns about the original shaft design. I've always hated the 2 piece design. From the beginning, I installed Toby Peterson's Bullet Proof Shaft coupler. It's really nice, and certainly does help with some of the issues, but there are still a lot more problems on the table. There is no coupler anyone could make that is going to fix all the issues with the design. The shaft has many reductions and neck downs to allow for all of the original splines to be rolled. This includes the splines at the coupler, which neck the shaft down to 17mm diameter (.66in). I never liked the idea of putting a lot of power through that small of a cross section. Even now, I do certainly push my car pretty hard, but I cannot honestly say that I've ever gone all out on it. Shown here is the original 2-piece shaft design:
Old Shaft.jpg

In addition to the shaft reductions, the splines themselves can simply wear out within the original coupler. With no lubrication, fretting corrosion takes place, slop builds up in the splines, and eventually they tear themselves apart. That is what happened to this shaft here, from a stock DeLorean with no modifications:
Broken 4.jpgBroken 1.jpgBroken 2.jpgBroken 3.jpg

For reference, this is what the stock shaft normally looks like (the splined end isn't broken):
103056.JPG

Sure there are single piece shafts out there made by GTO Racing/Quaife, as well as the now defunct Holloway Performance, but the price was never something I could justify. The exchange rate fluctuates but it pretty much always hangs around $5,000 for that gear shaft. At the end of the day, $5,000 was more than my engine swap cost, so I just couldn't do it. In addition, I never liked the idea about changing all the gear ratios. The GTO Shaft lowers 1st and 2nd to make them longer, primarily for the Lotus Esprit, and other UN1 boxes. The Lotus Esprit and the majority of other boxes have a 3.89:1 (some are even 4:11) differential instead of 3.44:1 like what the DeLorean has. Given that we already have the lower differential, changing the gear ratios to be even lower always seemed unnecessary to me (and gave up some actual wheel torque).

Last April when Car Fix asked me to be on their TV show to increase the power, and they started saying things like "500hp at the crank shaft," I started getting really really nervous. They even talk about it in the show. Thats why on the dyno, Jared the host said that we weren't going to push it to try and get that big number, because of the concerns about the transmission. Once I knew the episode, along with those performance upgrades were happening, I knew I had to do something. That's when I started modeling, and eventually drawing ...
New Shaft.jpg

I decided that since there was no shaft out there that fit my needs and my budget, I'd design my own. It took almost 8 months to design, find a shop that could do the work, get the tooling, and have the first one fabricated. There is a whole hell of a lot to learn when it comes to gears, especially when it comes to metric helical gears. There were a lot of times where I really wasn't sure if it was going to happen. Still, against the odds, earlier this week my first shaft was delivered!! I spared no expense or engineering opportunity when it came to material used, machining process, heat treatment, and post processing. The end result came out better than I could have imagined. Simply from the look of it, the strength increase is obvious. Shown here is the cross section of my new input shaft, compared to what the stock setup looks like:
Original Shaft Assembly Cross Section.jpgNew Shaft Cross Section.jpg

Here are the pictures of the first shaft, along with my huge book of transmission research (assembled over the last 3 years or so), some of the tools (24" vernier and vernier protractor not shown), as well as some of the rapid prototypes I had made to test my gear profiles:
Shaft 1.jpgShaft 3.jpgShaft 2.jpg

The shaft is a single piece, designed and machined to be a direct drop into the Renault 369-08 transmission that is in the DeLorean (a hand full of very early cars had the 369-03 which had the different input shaft spline pattern, though this shaft would work if you simply use the standard clutch). No extra specialty tools or modifications are required to install this shaft into a DeLorean manual transmission. The gears are all designed to mesh with the existing DeLorean gears. I did my first test fit of all of the parts and gears assembled on the shaft and everything lines up perfectly! All the gears mesh very well and everything fits just as planned!!
Shaft 5.jpgShaft 4.jpg

My goal is to get this shaft installed into my car (unfortunately it probably wont be until after the new year) and then begin testing. By testing, I mean I'm going to drive very very fast, with wide open throttle I've been collecting transmissions and now have about 2-3/4 extra transmissions to play around with so I'm not afraid to push it as hard as I can.

Once I'm comfortable with the design, I'm considering making more if there is interest shown. I'll have to really gauge it though since some people are interested in actually buying, where others are just interested in having it available and on the market. I'm not someone that is okay with taking money for pre-orders because of how hairy that can get when it comes to delays, problems, etc. If I'm able to do a run of them, it will be based on what I can afford to have made. Anyway, all that logistical stuff can come later. For now I'm just excited to have the one!!!

For those curious, the shaft is AISI-4340 that is through hardened and tempered with some post machining after heat treat. All of the gears are machined by hobbing. The splines for the clutch and the splines for the 5th gear dog teeth were also machined by hobbing. The 3rd/4th splines are milled and indexed. Both helical gears (1st and 2nd) had their roots and faces shot peened for additional strength. I am also considering having the shaft cryogenically treated for added uniformity and strength.

[Farrar]

[David T]

It isn't just about going fast. A lot of it is about shock loading. For instance dropping the clutch, spinning the wheels and when they grab the whole system must absorb the shock loading. Because of the length of the input shaft it must be able to twist and not break. You need the right combination of materiel and heat treat to make it ductile enough to twist but strong enough not to break. It is not all about cross-sectional area but as you pointed out that is important too. Every transition becomes a stress concentration point so the "trick" is to make them as gradual as possible. Then there is the logic of having a "fuse" or weak link somewhere in the system by design. Hopefully a part that is easily and cheaply replaceable and it's failure won't cause a cascade failure of other parts. Modeling can only get you so far, after that you must do testing. You are now doing what Legend Industries was up to when they increased the HP of the motor with twin turbos and found they were wrecking input shafts. As you also noted the same input shaft is used in the Lotus. My suggestion is to increase the available market and offer a drop-in replacement to the Lotus community also. I wish you luck but this can be a very expensive endevour.
David Teitelbaum

[Nicholas R]

It isn't just about going fast. A lot of it is about shock loading. For instance dropping the clutch, spinning the wheels and when they grab the whole system must absorb the shock loading. Because of the length of the input shaft it must be able to twist and not break. You need the right combination of materiel and heat treat to make it ductile enough to twist but strong enough not to break.

I agree about shock loading. Thsi was the logic behind going with the 4340. I went back and forth between 8620 and 4340 but 8620 is really good for gears because of it's ability to case harden, but 4340 is higher strength, and is better for shafts. In this case, I had more faith in the design of the gears (they're already a pretty robust design) than I did the shaft. Dont get me wrong, the 4340 is also great for the gears, but it is excellent for the shaft. I found an excellent balance between the strength and izod curve on the 4340 hardening graph that I believe will give excellent results.

It is not all about cross-sectional area but as you pointed out that is important too. Every transition becomes a stress concentration point so the "trick" is to make them as gradual as possible.

I agree, avoiding any kind of neck downs or notches was really important to me. I tried to keep the shaft as consistent as possible without adding any features that would add a "notch factor". The lowest cross sectional area is about 23mm where the first snap ring goes, as opposed to 17mm on the original shaft.

Then there is the logic of having a "fuse" or weak link somewhere in the system by design. Hopefully a part that is easily and cheaply replaceable and it's failure won't cause a cascade failure of other parts. Modeling can only get you so far, after that you must do testing.

I have to be honest, I disagree somewhat with this. It has only ever been on this forum where I've heard people talk about the coupler being a "mechanical fuse". Never on the Lotus, Renault, or GT40 forums have I ever heard someone speak as though the coupler was needed. I agree that in theory this isn't necessarily a bad idea (in some cases), but I dont believe that this is the purpose of this coupler. Truthfully I think the coupler is there for 3 easy reasons:
1. Ease of machining - It's a lot easier to machine two 12" shafts than it is to machine one 24" shaft, espeically when one shaft is simply 2 sets of rolled splines.
2. Ease of assembly - It's easier to assemble the entire gear train on a 12" shaft then it is on a 24" shaft, especially when you're trying to do it at a production rate.
3. Versatility - The Renault UN1 and 369's are both decendants of the 367, which was used in different vehicles that required different length clutch shafts. The 2 piece design easily allows for different length clutch shafts installed. The line also includes the UN5 which has a different clutch shaft than the UN1. Not sure about the UN7, haven't done much research on that.

Plus, look at the picture I posted above of the failed input shaft. It failed because the teeth inside the coupler stripped out, but look at what broke; it wasn't the coupler but instead it was the main gear shaft. The splines failed, so the entire load was going through the roll pin, which simply sheared off the end of the main shaft. Like you said, it was a cascade of failures. That transmission is essentially useless now unless that main shaft is replaced. In my opinion, the coupler in this transmission serves no valuable purpose.

You are now doing what Legend Industries was up to when they increased the HP of the motor with twin turbos and found they were wrecking input shafts. As you also noted the same input shaft is used in the Lotus. My suggestion is to increase the available market and offer a drop-in replacement to the Lotus community also. I wish you luck but this can be a very expensive endevour.
David Teitelbaum

I've thought about trying to expand the market. The difficulty is that the shaft will only work in gearboxes that have the same gear ratios as the DeLoreans (the Renault 369-0. First gear has to be 3.36:1 and second gear has to be 2.06:1. The good thing is that there are a lot of other UN1 and 369 boxes that share those ratios so hopefully it will be possible. It will just be up to the customer and myself to ensure that they have the ratios they think they have. Thanks for the luck!! I've already discovered how expense this can be....

[Josh]

Amazing work Nick.

[BABIS]

awesome work! now you need just a new LSD

[ccurzio]

Very, very nice work.

If this were to become a production item, do you have ANY guess as to what an individual unit would cost?

[David T]

Very, very nice work.

If this were to become a production item, do you have ANY guess as to what an individual unit would cost?

I agree the coupler was never designed to be the fuse. It may just have worked out that way and was a convienient way to modify the transaxle to couple to the PRV. Snap ring grooves are a very high stress concentrator and should be avoided whenever possible. So are holes. Too bad things like this are unavoidable.

[Nicholas R]

awesome work! now you need just a new LSD

It's definitely on my list!! My goal is to eventually create the ultimate DeLorean manual transmission using the single piece shaft, a Quaife Limited Slip Differential, an new clutch from either Centerforce or AP Racing, and a transmission cooler (that I'm also slowly working on).

Very, very nice work.

If this were to become a production item, do you have ANY guess as to what an individual unit would cost?

Tough to say. The tooling and process development for this first one has been pretty expensive. I can't imagine I'll be able to afford more than runs of 3-5 shafts at a time (and even that could be pushing it). At least for now. If I can bring these to the Delorean market, my goal would be to try to get the price to be at least half the cost of the existing shafts that are available (essentially shooting for the $2000-25000 range.) I know that's not cheap but it's better than $5000. I wish I could say that this machining was cheap but it's not; at least not at the quality I'm looking for. Especially when you add all the heat treatment and post processing.

Time will tell. I'll start with small runs so that if there ends up being no interest, hopefully I won't be out too much.

[Nicholas R]

If I can bring these to the Delorean market, my goal would be to try to get the price to be at least half the cost of the existing shafts that are available (essentially shooting for the $2000-25000 range.)

I meant for that to be $2000-$2500 not $25000, haha. Hopefully I'll have a better idea in the next few weeks/months, right now it's way to soon to say.