scathcart

I've done much of my own independant testing on these motors: the RA test was my first, and I have done 1 more on RA seals since then. I have done testing on 2 motors using Atkins seals since I began testing. My first RA testing was done in car... for testing purposes, it was too much of a hassle.

Instead of installing the engines into a car as before, I built a jig to mount the engines to and then rigged up a spare stock ECU to run the engines. Emissions equipment was removed from the engines in testing, all of the rest was stock, save an adjustable FPR and a massively ported wastegate. Fuel pressure was set to a base of 40 psi.
To the flywheel of the engine, I bolted a large brake rotor using some steel 3/8" brackets to attach it and centered it with a dial indicator, and to the bottom of the jig, I welded on a bracket to attach a brake caliper. I ran some custom steel hard line from the caliper to a 3000 psi electric hydraulic pump, and filled the system with brake fluid. I tee'd into the hydraulic line and attached a 3000 psi gauge... Basically, I creating an engine brake to produce load on the engine so that I could produce boost.

For exhaust, I mounted a 3" downpipe, with a FJO wideband into it, datalogging to a cheap laptop. I used a lawn mower choke cable to control the throttle. I used a stock radiator to cool the engine, with a 16,000 CFM 48" shop fan blowing through it. (yes, 16,000 CFM) I used some PVC piping and relocated a stock intercooler to in front of the shop fan to keep intake temps down.

I would run the engines for a break in of two hours at a 3000 rpm under 100 psi of brake pressure. Exhaust was collected into a piece of rubber pipe, which led to outside my shop door. After the break-in, I would do some mild testing to determine how much brake pressure was required to hold the engine steady at WOT at 6000 rpm.... this was tricky, as I did it all ducking behind a stack of 6 sheets of 3/4 plywood behind the flywheel (which I am thankful for... explain later), with my pump, laptop (for rpm) and throttle cable in-hand.

Once I was able to stabilize the engine at ~ 6000 rpm WOT, I would cut the rear secondary injector... and the engine would audibly ping, drop in rpms, and then stall from the excessive load on an engine now running on one rotor. At this point, I would let the engine cool down, and dissassemble. For anyone wondering, AFR's were at ~10.5:1-10.8:1 at WOT at 6000 rpm at 6.5 psi, and spiked to 19:1 or higher when I cut fuel.

My first test using this particular set-up was a set of atkins seals. At 6.5 psi and my induced detonation, 2 of the atkins seals let go. No other seals were damaged, and the endplate was fine. Both apex seals broke where they meet the corner piece, and the seals stayed in their grooves.

I rebuilt the engine again, reusing the coolant seals, and installed a new set of RA apex seals in the rear rotor. Under the same conditions, at 6.5 psi, the rear corner seals all broke, one side seal broke, and the rear iron cracked. ALL of the RA apex seals remained intact. This engine did not stall nearly as fast as the others.

The next day, I replaced the rear apex seals and the rear iron (again with atkins seals) reassembled, and tested again, at 12 psi, via a cheap ball-and-spring manual boost controller. Same break-in, same induced detonation, and all of the atkins apex seals broke.... I assume they broke in the middle, as the housings were extremely scarred directly above the exhaust port. From this, I deduct that they broke in the middle, and then flexed out from centrifugal force over the exhaust port but remained in position due to cocking against the rotor apex seal groove and the edge of the exhaust port, and then crushed themselves against the top of the exhaust port, and crushed. The rear rotor was also ruined.
On this test, one corner seal broke, and the rear iron remained intact.

I took the engine apart again... reused the coolant seals, replaced the rotor with a spare S4 N/A rotor I had laying around, and filed down the grooves in the rotor housing with some sandpaper (this didn't need to be a high-quality engine, and grooves above the exhaust port are not crucial to engine compression).
For kicks, I slid in stock apex seals, turned up the fuel pressure to 44 psi, and set the boost controller to 15 psi.

This testing... was not pleasing... from what I can tell from the datalog, when **** went wrong, I was at 5500 easing off the brake pressure slightly to bring the rpms up. From what I can tell from looking at the mangled mess of iron, one of the welds (AC stick welds) on the jig on the exhaust-side engine mount let go... and the jig was bolted to the floor. The entire engine set-up went catapulting over to the side of the shop, and took out the shop wall (This is why I was glad I was behind the plywood barrier).

The radiator, oil cooler, and intercooler had also been bolted to the floor... the radiator was sheared in half, as was an oil cooler line, which made a HUGE mess of everything. The wideband went for a bit of a ride (killed the sensor, but the controller was fine), the throttle cable was ripped out of my hand, and the hydraulic line sheared off. The UIM physically cracked off of the UIM, and fuel lines were ripped off, spouting gasoline out all over the floor (as I had stupidly wired the fuel pump to 12V constant).

After about 30 seconds, I figured out fuel was dumping, and ran over and kicked the battery cables off the battery. I grabbed the fire extinguisher and held it in hand, while I dumped a garbage can of kitty litter onto the engine-mangled mess.

Figuring I had staved off death enough and with my jig destroyed, I concluded my testing. If I ever choose to do this again, it will be safely on a water brake dyno.

The only real loss from the engine fiasco was I had to replace a wall stud and a piece of plywood in my shop, a wide band sensor, and all of the crappy parts from my test engine. The test engine was built using parts that I found too risky too use on any customers car, so it was no big loss.





To sum it all up: atkins seals broke and tore up housings where RA seals did not. Its FAR from perfect testing conditions, but its all I've got to compare by, and likely the same for others.

RETed

HOLY CRAP!
You got a deathwish or something???
Glad you're safe!

Basically, what your experiement concluded is the Atkins apex seals are softer than the RA apex seals.
Which most people have already concluded.

BTW, did you know how the Atkins apex seals broke in the center?
Was it all chewed up or did it have a fairly straight crack down the center?


-Ted

Parastie

Jesus christ Scath, are you insane? LoL.
Man, too bad you didn't tape that, i would actually PAY to see that video!

scathcart

I put more trust in my welds then I should? I was excited and in a bit of a rush when I did it... doing it again, I would have TIG'd it and taken my time.

When the atkins seals broke on the second test, they bridg3ed out the exhaust port and caught on the closing edge of the exhaust port. I had placed a "apex seal collection screen" in between the exhaust manifold and the turbo... basically just a very loose-screen mesh that was ceramic coated. Unfortunately, the pieces contacting the rotor housing were too chewed up on the rotor housing to analyze.

When the atkins seals cracked near the corner-piece, they had a odd chewed texture, not like a crack that you would expect from stock apex seals, or the RA seals. It almost looked like metal does when you bend it back and forth until it breaks.

Less carbon in the atkins seals and lack of hardening is likely the cause of this.


From all of my testing, I have found that: RA seals are the best for standing through detonation.
Stock, 3-piece apex seals are best for sealing and power (outside of ceramic apex seals).

scathcart

In a recent weekend, I got bored, and I brazed horseshoe nails together (they are flat nails) to form throwing stars, which would stick into the shop wall. I put a rag on one of them, soaked it in carb cleaner, and then threw that at the wall... which, of course, caught on fire.

I always keep three fire extinguishers in the shop at all times, lol. I use them a lot.

Rob B.

Wow, I was looking at prices for rebuilding my T-II when I get it out of the car, and the RA option just keeps looking better and better. Now, the main difference I saw between RA's basic set and Atkins type B was that the RA kit didnt come with corner seals, side seals, rotor oil seal sets and an oil pump chain. What is the likelyhood of these parts going out? (my engine has 187k, I believe, it could have been rebuilt before, but I doubt it.) And do they justify Atkins roughly $500 premium over the RA kit? Thanks
-Rob

scathcart

Many people don't replace the corner seals or side seals, but I always fit new ones.
Oil chain and the hard oil control rings are usually fine to reuse.

Rob B.

How would I go about checking to see if I needed new side and corner seals or a new oil pump chain and such? I really want a bulletproof motor before I even think about upgrading it. Thanks
-Rob

RETed

The metal corner seal part does not usually wear significantly.
I highly recommend to change the rubber center section, which can be purchased separately from any Mazda dealer.
RA also sells this (dunno separately), but it's a piece of "cord" that you cut to length to replace the old pieces - it's a bit hard to explain in text, but once you get it it's self-explanatory.

Oil chain slack specs are in the Mazda FSM.
I don't usually replace it in most cases unless customers insist on it.

Side seals very rarely wear significantly.
Grinding down new side seals from Mazda is a royal pain in the ***.
It's best to reuse the original ones (keep the positions marked!) for newbies, as this keeps the factory clearances.


-Ted

Dave777

Speaking of Atkins softer seals, I contacted them in mid '04 for apex seals on a rebuild I did and they told me they had new "harder" apex seals they were just starting to get in. Anyone else heard of that? I havn't noticed anyone talking about this, but that's what they said.