Rally Revival: 1973 Mazda RX3 Rally Car

Shavarsh said:

Agreed 8.8 is hard to beat on paper. I put one in my pinto (shortened explorer), ran it for years, and eventually replaced it with miata irs (with 8.8 diff) because of terrible ride quality (lots of extraneous factors could have caused the poor ride though). I remember looking at Ford 7.5 and Toyota small truck axles (like a mini 9") among others. Obviously the 8.8 won at the time though. YMMV

 

The devil you know (FC parts) sounds like a great way to go from here. Unless part of the fun is a new learning curve.

I think with this project I am going to need to pick my battles and farm out some others. The learning curve I'm aiming on this one is improve on my own fabrication competency and maybe some new toys (tools).

Updates when I can get inspired and find the time but that brings us to the next round of fun! I managed to sell off some parts from the dashboard area that are rx3 specific and random bits and pieces that wont be needed for the resto-mod, doesn't hurt to have an extra 500 bucks in your pocket for basically trash!

Trying a little new school meets old school. 

enter the 3d scanning. I wanted to learn more about this tech and understand it better and its tied to a "work" project so this is a good way for me to get exposure of the process while also using it for my own benefit. Its just a simple (older) Ipad Pro w/ LiDAR, the phone LiDAR would work too as I don't need this big screen to scan tbh

don't mind the reflection but it was pretty spot on for double checking measurements

the idea would be to build the tubes in CAD and get them pre-bent/notched to save on the fab work and equipment but it sounds easier than it is. much like when someone says "just 3d print it!" its far more involved than just clicking the go button - which is what i have just learned from 3d scanning, often times I have said "just scan it" but not realizing the steps it takes to get the detail/data you need... Maybe more to come here, maybe not. Definitely worth the hands on trial

so luckily this car is constructed fairly simply so time to start cutting away at it and adding structure back to it and build from there:

(the inner fender is just 1 piece of metal, the rx7 is 3-5)

and time for the exploratory cutting 

this is good practice for the new welder and get some beads under my belt before I get to work on the rx7 repairs. since that is on hold at the moment (in que) I can work a little bit on this and see if its still a bad idea or not. 10 bucks in scrap metal and I can fix these cab corners and tie them to the rockers then start the base for the tube front end from there. 

some more rust removal and viola! 

I'll do this on both sides as they have similar rust holes forming. .125" wall may be overkill but I didn't see any .095" anywhere for angle iron. 

now considering powertrain options is where the scan packaging data will come into play @nonack had coined this great "its N64 goldeneye graphics" so hard to reverse engineer or fine detail things but for general space and packaging this can work. however I have to play by the rules so this will be tight:

I can do a 3" widebody but thats based on the internet measurements. I can put some big honking mirrors on them and call that the outside and measure from there and add 10% but all signs are telling me with some minor flares I could fit whatever I'm looking to swap under the body no problem. Frame rails are junk and will be re-created so they possibly can be moved too.

  more to come

Those numbers above are from the internet and various specs online but If I measure from the width of the outmost  of the mirrors it is  65.5" across. Which means, with that width I have an allowable max width addition of 6.5 so 72" in total. Not a big difference but any inch counts here because its so narrow. 

For example, the width of an 05-07 STI (with gravel rally tires) from outboard to outboard is 69" so this is feasible to hodge podge with an array of parts:

enter the digital (augmented reality) engine swaps:

No I don't want a subaru 6 cyl in the engine bay but to understand where the shock towers are to correlation with the trans mounting flange

its not good enough to make a digital twin and start working in CAD but it works for general swag which is nice a real scanner would come in handy here but I lack all the other downstream skills to make that work and that aspect is out of budget/scope at this moment. 

speaking of spreadsheets, the other problem to this is no torque and non turbo so I'm stuck with 3.90 or 4.11s with a subaru trans, a 4.444 would be ideal with the DCCDDCDCCDDDCDDDCDs center diff magic but still its a LOT of gear for 250hp (goal) with 2500lbs.

(5 speed WRX gearbox, 13b shift points)

50mph in first gear is way too much much like it was 45mph with the FC setups I've done before.

 the 6 speed is what I need and 4.444 but availability and pricing is going to be a bit wild, and basically for rally the OEM open style/viscous is not going to work well so the playing field narrows for 6 speed, STI, 4.444 ratios and DCCD and also doesnt exist so I need to 4.444 swap a 3.90 ratio 6 speed and if i'm in there I would be adding an LSD so quite an expensive gearbox...

which is much more in line with racing needs for rally and current setup: (rx8 gen 2 transmission and mazda competition rear end) the 4.444 on the 5 speed is only 3-5mph difference per gear so hardly noticeable vs the 4.11 in this space. 

woof

EDIT: after consulting the subaru wizards that I know; I now know that the 6 speed STI trans never came in 4.444 R&P ratio but for aftermarket it is available and pricey.  so you can't mix and match R&P between the 5 and 6 speeds for the transmission. 

How did I not see this thread before?

Let me tell you what I was going to do with my RX-3.  (It was rustier than yours, and a 4 door so panels and glass was even harder to find)

I did some measuring at the Clapped Out Garage where Volvos roam wild and free, and found that Volvo 740 front parts looked like they would slot right in similar to how an FC crossmember fits an FB with some very minor bracketmaking.

The struts are about the same height, only a couple inches narrower at the towers (so you get a little negative camber), the crossmember would sit about where the RX-3 one does and is the same shape, and the real kicker is that it has rack and pinion steering mounted INSIDE the crossmember.  It looked like the steering shaft would clear a rotary just fine, too.  The only issue is having to make rear mounts for the control arms, but if you are making tubular arms then you just make a bracket for rod ends like the compression rods MkII Escrot guys use to decouple the sway bar from the TCAs.

For the rear I wanted to do a Escrot style 4 link using a Ford 7.5" from a Ranger, using two short axles from a Mustang.

I had a way, in theory, to have a bolt up solution for 4x4.25 front and rear.  I proceeded to buy a set of 4x108 Compomotive gravel wheels and then realized that the rust was beyond my ability to repair.  One "oops" removing the windshield to start repairs and the car would be totaled.

The nice thing about the Volvo parts is they use a Miata style spindle and not loose bearings like FB/FC.  And in theory rally struts are available since this is rallied in Scandinavia.

That gas tank is stock.

When Mazda made the RX-3 they took a long hard look at the Mk1 Escrot and copied a lot of it.

In reply to Pete. (l33t FS) :

haha I have no idea, I know we have pondered similar ideas before 

I've only just begun the, add real metal to air portion by bracing and fixing the firewall/cab corners. if I get the firewall and rockers sorted I'll start building the engine bay and likely the engine bay build part will be a 2026 part. maybe get it ready for a car rotisserie? I don't want to weld upside down that's for sure.  if I have a tube front end I can put whatever the strut towers almost anywhere I want (gain camber and caster).

I'm still gonna mull over AWD or not. it would be easy for me to just keep it RWD and enjoy but this part I can work on the adding basic structure so I can pick it up on the lift. 

need to work on welding better but thick to thin sucks and thick to rusty thin is worse. 

the first weld on there was fantastic, then the rest were E36 M3. so back to grinding, cleaning and re-trying. 

Even tying the roll cage in, those pieces of angle iron are way overkill, IMHO. But since they are already started, probably leave em, unless you really wanna cut ounces to make pounds down the road. As for Welding them in, to connect thin (crappy) metal to a thicker piece, First, Clean clean clean! Then, I usually run a mostly superficial bead along the thin stuff and right next to the thicker metal if you're just adding a bit of steel and not trying to actually tie the 2 together, you'll avoid burn through. Then you can go back over and add a bit more heat welding in the thicker stuff and likely get better results from the now thicker, cleaner metal. Hope any of that helps!

I could only find 1/16th in aluminum but 10 bucks for these two pieces from my local metal providers scrap bin was too hard to pass up. they are a little long but will get trimmed later. plus I can use this space to put a bar across and keep some rigidity while fabrication occurs and cut away slowly at the rot. I got better about moving quicker and getting the thick to thin rather than thin to thick and it came out better but for problem areas I'll give that a try.

I'm also going to use this spot to temporarily attach the cage that is in there to keep the rest of the rigidity. might be overkill but a couple welds and some sacrificial tubes to prevent the car twisting enough to break the windshield is good enough for me.

Vintage Rally, WOOT!!!