Involuntary Engine Rebuild

 I've always wanted to rebuild a rotary engine.  I imagined I would determine the timing and I guess I did, in a way.  By melting my coolant seals and perhaps cracking the block or other miscellaneous mischief.  So, this winter is rebuild winter.  Cruise control is again delayed.  There may be some advantages however:

1. I can shorten up the incredibly long wires on my oxygen sensor.  That thing could sense oxygen to the back of a semi.

2. I can paint my radiator mounts and lower engine mount.  They were last minute and I was burning to get on the road.

3. I can rewrap my wiring which had several runs going outside the harness.

4. I have a new clutch disc which will reduce my pedal effort and expand the modulation zone.

5. I can plug the open vacuum line on the intake manifold that I found.  That must have been part of the idle stability problem.

6. I can reflect on the need to shut the engine off directly as it overheats!

As usual, this project did not go as I imagined.  I thought the easy part would be engine removal.  Wrong.  I really wedged that guy in there!  There are only five bolts holding the bellhousing to the engine.  Two are easily accessible from below.  The other three, in my very unique case, are just underneath the firewall!  One I was able to break loose with a normal wrench.  The others I couldn't reach.  I couldn't get fingers on the loose one to actually back it out and remove it.  So I bought this:

I don't normally advocate specific products but, in this case, everyone needs to buy this remarkable thing!  That wrench ratchets on like 1.5 degrees and has a small profile so I could break the other two loose.  Then, I finally realized, I could remove the engine mounts and lower the whole thing until I could remove the bolts.  That's a paragraph to describe about 6 weeks of short attempts interrupted by holidays and business travel!

In any case, "We ain't never not done it yet!"  Below, the engine emerges!

Didn't even have to drop the front suspension clip.  Here it is on the floor:

I elected not to use an engine stand because I wasn't sure where to support it.  I had to buy a special set of tools for flywheel removal:

Shout out to Racing Beat for those.  Not cheap but make short work of removing a 350lbft torque nut and popping the flywheel off of its taper.  By short, I mean 10 minutes.  The 54mm socket alone would probably have been $40 and I don't have anything long enough to drive it with.  I guess that means it is cheap!  Easy part.

Having the engine out, I began to struggle to tear it down.  It's simple in principle.  A dozen or so long bolts hold the stack of castings together.  However, should one remove only 11 of the 12, it's very difficult to pry the stack apart.  I actually fabricated a jack screw arrangement to push the plates apart only to discover number 12 hiding in a recess in the casting!  Damn.  Came apart easily after that.

Now the diagnosis.  There is a relevant line from the original X-men movie.  Storm is about to best the evil mutant Toad.  She calls up a thunderstorm and asks him "Do you know what happens to a toad when it get's struck by lightning?"  He can only shudder and she finishes "The same thing that happens to everything else!"...and fries him.

I could just as well ask "Do you know what happens to rubber seals when you expose them to combustion?"  You guessed it: "The same thing that happens to everything else!"  And I fried those:

Looks like I may have melted some aluminum there as well.  Had I been able to imagine what was happening, I would have stopped and called a tow truck.  I thought driving at low throttle and frequently refilling would be fine but alas, no.

This is an amazing engine though.  Invented in the 1950's in Germany and brought to production in Japan, how they worked out the kinematics and machined the double trochoid without CNC is beyond me.  On the other hand, I feel like we had a better understanding of how things worked before we could simulate them.  Back in the day, they connected a monometer to a rotary valve with 30 openings to the combustion chamber.  They then lined up 30 monometers, one for each opening.  You then had a live graph of combustion over crank angle.  You could make pencil marks or a picture at different speeds and throttle positions.  

Of course we can now make that graph with resolution to 0.01degrees accuracy and print it in color and simulate it and print that in color.  For sure there is value there, but I'm not sure that someone sitting at a PC all day has the same understanding as the person who had to connect 30 passages to the monometers and knew each crank angle.  Then again, I am old.  Hence the name of this blog!

To be honest, I hadn't truly understood how one spins a triangle around an oval and assures a seal at all positions.  Now it's clear.  It's a spirograph.  The eccentric shaft (I still say crank shaft having grown up with piston engines) drives the rotor around an eccentric orbit.  There is a fixed gear the other side of this picture that forces the rotor to follow the trochoidal path.  Like the whole engine concept, it's simple.  Sadly, for humans simple is not easy to see.  In fact we live to complicate things.

But let's talk about what I've done.  How much damage did I cause with the overheat?  For sure, I melted out the coolant seals.  

Those are melted rubber seals!  The rest of the damage, however, I cannot accept credit for.  I think I paid $2200 for a nearly worn out engine/trans combo.  I work for a clutch company.  We can still measure a clutch in the US!  Here's the graph:

The diaphragm spring clutch is quite an invention.  You can see the double hook load curve.  The linearly rising line is the lift-off of the pressure plate.  This allows you to shift when you depress the clutch pedal.  The remaining curve is the release load.  This is what you feel when you push the clutch pedal.  Before diaphragm springs, we used coil springs.  They have a linear characteristic.  So, as soon as the clutch disc wears, you lose clampload.  You start life at the highest pedal effort and go down from there and start slipping soon!  The diaphragm however actually gains load for the first half of the wear.  You can see the vertical line in the graph.  Where it intersects the double hook is the load applied to the friction material.  This creates the torque.  This clutch is near end of life so the vertical line is to the left.  It started life to the right of the peak.  As it wears, it moves left, first gaining clampload and then losing it as it crosses the peak.

This extended clutch life greatly.  In fact, the biggest innovation until we at Schaeffler (then LuK) invented and perfected the self-adjusting clutch, which doubled the wear reserve of even a diaphragm spring clutch!

It's not obvious in the graph but the clutch was only 5 or 10k miles away from replacement.  It was starting to disintegrate with fiberglass packed around the bolts.  Just in time!

My biggest question was "How much damage did I do by overheating?"  I checked the rotor housings and they were not warped at all.  Re-usable by the Mazda factory manual.  However, there was significant damage to the chrome in several spots.  This was likely building up over many years.  The engine was advertised at 60k miles and they don't seem to have been kind miles.  I am buying new housings.

But I feel better knowing I wasn't far from a rebuild, which I've wanted to do for years, in any case.

In future posts, I will amaze you with stories of assembling 30+ parts of rotor seals for each rotor and hopefully, getting it all back together again!  In the meantime, keep on having #funwithcars!