There are two fixes for worn rotor housings. Replace with new, as in this build, or have them honed. I have had good success using outside vendors but have also had some not so good outcomes so be very careful with whom you deal..
Here’s a pair of honed housing pictures:
If found not on spec an adjustment will need to be made to the spacer.
It is nice to have a few available as well as a 0 to one inch digital .0001 micrometer as well as a few spacers.. often to get the right clearance a bit of modification to a spacer is in order.
You will need this:
Built, fixtured and tuned correctly, your rotors will look this. They came out of my motor and that had been in service 4 years around the 500 rwhp level. Pumpgas and 100% methanol AI. They are pictured exactly as they were as removed, I E, not cleaned in any way.
So why does Mazda specify such a low torque spec considering the bolt is of very high strength? There are 6.3 inches of aluminum (rotor housings) in the stack and aluminum is second only to plutonium as to thermal expansion. While the tension bolt is not driven anywhere near it's Proof Load (75% of Yield) by the recommended torque, as the stack expands the bolt heads to that vicinity.
My conclusion is that it is unwise to attempt to solve the clamp issue by adding much torque to the factory tension bolts as tension will further increase towards the inelastic/yield area as the stack expands. A modest increase in the tensile strength, such as the Turblown studs should be helpful. Further, almost 11% additional clamp is added since the bolts are 10 mm replacing the factory 9.5.
Finally, buyers would be well served to be careful as to which stud kits should be added as two of the three tested did not deliver as much tensile strength as the factory tension bolts. One brand had only 51% of the yield strength as the OE bolts!
Building a set of rotors generally takes me about a full day. They could also be assembled in a couple of hours.
Patience is rewarded.
The target is enhanced compression and one of the keys is sideseal to cornerseal clearance. Sideseal clearance is similar to ring end-gap on a piston engine. My target is .002 on all 12. I am not happy with .003.
Ready to attack the oilpan
Approximately half of the front aluminum front covers have stripped front boltholes. Unfortunately this speaks to the level of some people venturing into the internals. Correct bolt torque is 100 INCH pounds. It is so easy to destroy the threads of a 6 mm aluminum bolt hole.
Once all the bolt holes are freshly tapped and all the mating surfaces are spotless, wiped with solvent, coated with HondaBond, and the pan is placed on the block we are not done.
All CPR motors have received a David Garfinkle original design oilpan/engine block brace. David was the first to realize the need to accommodate the raised stiffeners on the pan rail. The mating side is relieved so it delivers clamp pressure to the 95% of the rail surface rather than just the ribs which are maybe 5% of the rail area. I am delighted, now, to recommend what i consider to be the best designed version of David's brace. It is always nice to see new items for our 26 years old cars. Thanks Chip, at Chipsmotorsports, for a brace that is so nice i really want to put it under my pillow. It has a beautiful satiny finish, finished edges, and countersunk holes featuring stainless steel Torx bolts.... absolutely gorgeous and will be on all of my motors going forward.
Proper torque for the Chips brace is 150 inch pounds/10.9 foot pounds as the bolts are 10.9 grade. All other braces are 100 inch pounds/8.8 ft pounds as bolts are 8.8.
Here’s the tool you need to properly install a pan… an inch pound torque wrench.. It is also the tool needed for the many other nuts or bolts with a 10 mm head.
Thee Mazda tension bolts provided a surprise. Given the interplate movement i see on incoming motors and the modest (24-28) torque spec i had them as being around Grade 5. Not even close, as they registered 150,000 Yield and 160,000 Tensile.
Block preparation is probably more than an hour as all 18 pan boltholes need to be tapped. Don’t think so?
All CPR motors are accompanied by a Final Spec Sheet. Here is the sheet on this motor:
CPR ENGINE SPEC SHEET
OWNER XXX December 20, 2017
ROTOR HOUSINGS FRONT REAR
A WIDTH 3.1490 3.14952
B WIDTH 3.14908 3.14900
C WIDTH 3.14931 3.14912
D WIDTH 3.14925 3.14917
WIDTH VARIANCE .0003 .00052
MAZDA MAXIMUM ALLOWABLE WIDTH VARIANCE .0024
EDGE GROOVING: none none
STANDARD 0-1/32 inch
MAX WIDTH 3/32
MAX DEPTH .020
HONE NOT APPLICABLE AS NEW
ONE NEW, ONE LIKE NEW
THRUST WIDTH 3.14278 3.14455
LAND PROTECTION .0115 .0105
SIDE HOUSING & ROTOR
CLEARANCE (.004-.009) .00622 .00445
WEIGHT GRAMS 4328 4334
(50 GRAM MAX DIFF) 6
SIDE IRON HOUSINGS FRONT MID REAR
CPR EXCELLENT USED
WARPAGE LIMIT .0016 .
SIDE SEAL WEAR LIMIT .0039
SIDE SEAL OVERLAPPING OIL
SEAL WEAR LIMIT .0004
SIDE SEAL OUTSIDE OF OIL
SEAL WEAR .0039
OIL SEAL WEAR .0008
BRAND ROTARY ENGINEERING
SIZE 2 MM COATED
APEX 1 SEAL TO GROOVE AT .002 ALL SEALS
SEAL TO GROOVE (.0015-.004 LIMIT .006)
APEX SEAL SPRINGS ALL NEW MAZDA FD INCONEL
FREE HEIGHT SHORT
STD .130 LIMIT .067
FREE HEIGHT LONG
STD .246 LIMIT .181
SIDE SEAL ALL NEW MAZDA
CLEARANCE TO CORNER SEAL
(.002-.006 LIMIT .016)
CPR TARGET .002 ALL SEALS AT TIGHT .002- .003
SS TO GROOVE .0015
(.001-.003 LIMIT .004)
MINIMUM PROTRUSION .020
CORNER SEAL MAZDA NEW FD
OUTER DIAMETER (.4327-.4336)
MINIMUM PROTRUSION (.020)
CORNER SEAL SPRINGS MAZDA FD INCONEL
ROTOR OIL SEALS
O RING BRAND NEW VITON
WIDTH OIL SEAL LIP (.020) .01
MINIMUM PROTRUSION (.020) .020+
STATIONARY GEAR BOLTS 10.9 RATED ZINC COATED
(CPR REPLACED STATIONARY GEARS AND ADDED NEW MAIN BEARINGS)
OUTER 1 INNER 1 INNER 2 OUTER 2
MAIN SHELLS DIAMETER 1.69447 1.69425 1.6947 1.69472
MAIN JOURNAL DIAMETER 1.69057 1.69125 1.69150 1.69042
MAIN CLEARANCE .0039 .0030 .0032 .0043
(OUTER .0032-.0043 MAX .0051)
(INNER .0024-.0031 MAX .0043)
SHELL DIAMETER 2.91535 2.9152
JOURNAL DIAMETER 2.91215 2.9120
CLEARANCE .0032 .0032
(.0024-.0031 .0039 MAX)
RUNOUT (MAXIMUM .0047) .00025
ENDPLAY (STANDARD .0016-.0028 LIMIT .0035) .0016
E SHAFT SPACER CODE C
LOBE TO ROTOR .006
OUTER ROTOR TO PUMP BODY .007
OIL PRESSURE CONTROL SPRING NEW MAZDA
FREE HEIGHT 2.87 INCHES
OIL CHAIN EXCELLENT
ENGINE COOLANT SYSTEM INTEGRITY TEST PASSED 30 PSI
GARFINKLE OIL PAN BRACE
STUD KIT TORQUED TO 40 FT POUNDS
COMPRESSION TEST PASSED 120+
OIL PRESSURE TEST PASSED 80 PSI
This was an especially gratifying build as the important parts of the motor (rotor housings and rotors are NEW. In addition, Rotary Engine balanced and clearanced the rotating assembly as well as adding their premium apex seals.
Initial evidence of the build quality was the compression test which registered approx 125. This is impressive as literally it was produced from the first rotation of the completed assembly.
Block stability was buttressed by the addition of 16 chrome moly studs replacing the factory Grade 5 thrubolts. Additionally, the important stationary gear Grade 5 OE bolts were replaced by metric Grade 10.9 bolts adding 43% additional clamp force.
All 12 sideseals came in at .002 clearance to the cornerseals providing additional compression seal.
The entire rotating assembly was dynamically balanced and the rotors are a scant 6 grams apart. The rotors were side and face clearance so as to keep them from rotor housing contact at high power output. All new bearings support the rotors and crank.
I recommend 1 ounce of 2 cycle oil per gallon as premix, NGK R7420-10 spark plugs gapped at .0023. A proper Auxiliary Injection system is mandatory for anti detonation and clean motor internals.
I am almost always available for you and encourage your calls on any matter regarding your FD.
Best Wishes to my partner in this project,
Here are most of the components that hide behind the front cover.
If the endplay is on spec (approx .0018) the process takes 30/45 minutes and among other things involves torquing the front crank bolt to 192 ft pounds, my six special six stationary gear 10.9 bolts to 25 pounds, mounting the front counterweight, oil pump drive sprocket and distributor drive sprocket, front hub and bolt. I wouldn’t be surprised if this process was skipped by some.
Here’s some of the hardware on the front end:
Leaky oilpans litter the FD landscape and with good reason:
Unlike other engines, the FD static oil level is ABOVE the oilpan to block connection!
If this isn’t challenge enough, most oilpans have been previously removed and are bent from the removal
Various versions of Permatex silicone sealers are used which degrade significantly in the presence of gasoline.
All 18 bolt holes need to be cleaned with a tap
An inch pound torque wrench should be used.
It is close to impossible to do it properly with the engine in the car
The silicone needs at least 5 days around 70 degrees to fully cure.
A proper oilpan brace is almost essential.
A proper oilpan install takes me two to three hours.
After a thorough degreasing/cleaning the oilpan is placed in my pan frame. The pan rails are cleaned with a ½ inch drill and 5 inch wire wheel. Here we are in the middle of the process.
In addition to confirming parallel, when housing width is known and combined with rotor width, the proper rotor can be fitted to the proper housing. It is not unusual to find a clearance advantage by swapping rotors front to rear.
Rotors are often measured for thrust width so as to make sure the proper amount of clearance exists within the housing but also should be measured for land protection. I recently had a set of rotors where there was insufficient clearance between the thrust surface and the side of the rotor.
If insufficient clearance doesn’t exist the side of the rotors’ tips will contact the iron.
Yes, in order to properly build a motor one thousandth of an inch is the order of the day… rotor housings, rotors, crankshaft journals and runout, oilpump clearances. Four accurate digital micrometers required.
The stationary gear assembly locates the main crankshaft bearings and indexes the rotors on the crank. The OE bolts are Grade 5 and entirely adequate for the stock 215 rwhp power level. Given that we are increasing power 3X, Grade 5 has no business being anywhere near the stationary gears. I replace the 12 stationary gear bolts with Grade 8/Metric 10.9. Clamp force increases 43%.
A close inspection of the four dowel pins will reveal one end with a short wear length and one end with a bit longer wear length. I suggest you flip them to even stress.
I use two sealants. Hylomar behind all the coolant rings and dowel pin O rings and HondaBond for the oilpan, rotor housing legs etc. I did a silicone study and compared 4 leading silicone sealers. I cured them for a week and then immersed them in gasoline and E85. I then tested them with my Durometer for hardness.
The front bolt has Loctite, Hondabond and lube on it prior to being torqued to 192 ft pounds.
SEE NEW SECTION:
TENSION BOLT TECH
Let’s start a build:
This particular 13B-REW motor is replacing a three rotor 20B transplant. The car is a beauty and will be in car shows as well as on track at Brainerd International Raceway. The motor is being built for 600 rwhp and will be serviced by a Borg Warner EFR 9180.
Here’s the car:
Another build failure waiting to happen:
Note the garfed pilot bearing seal. The small breech caused by a miss when installing the transmission would eventually drain the grease from the roller bearings and lead to failure which would require transmission removal.
It is worth noting that I commonly see incoming motors with sideseal clearance miles from spec. I often see clearances in the area of one tenth of an inch resulting in significant compression loss and pressurized oilpan windage which results in a fast filling overflow tank.
There are a number of decisions to make re rotor hardware.
Corner seal springs are more important than you might think and there are options.
They are the first component to malfunction in response to detonation. Detonation raises combustion chamber temperature (CCT) from 1400 F to around 3500 F (not a misprint). Corner seal springs are flimsy items stamped out of thin metal sheeting.
When I lost my fuel pump fuse at the 2013 Texas Mile my engine went lean and Combustion Chamber Pressure and Heat skyrocketed. I was unable to restart the motor after driving to my paddock.. A compression check showed the motor down 30%. Upon disassembly the only item on the motor that was off spec were the corner seal springs which were flattened.
Hondabond blew the other 6 away.
Hondabond is made by ThreeBond, a global company headquartered in Cincinnati, Ohio. Threebond makes Yamahbond, Mazdabond and a bunch of others as well as other Threebond Silicone sealers none of which are the same as Hondabond.
As a part of my initial spec I check the crank for runout/straightness:
A dial indicator delivers the clearance verdict.
Next up... setting crankshaft endplay.
Finally I wrap the exterior of the pan rails to lower block with Hondabond.
The last step is a trip to the compression table. Both compression and oil pressure are checked. I really love my Rotary Compression Tester not only for its ease of use but also because it corrects for RPM (ESSENTIAL) and altitude. Compression checked in around 120+. Oil pressure 80 pounds.
CPR motors make 80 PSI oil pressure at 8000 rpm.
While not absolutely necessary at 600 hp the motor was dynamically balanced and the rotors were clearanced as to width and face along with trimmed tips..
A set of accurate digital micrometers are necessary to build a rotary.
This motor will use Rotary Engine coated apex seals, Mazda corner seals (I like the chrome exterior for lower drag), FD OE inconel corner seal, apex seal and sideseal springs, along with Atkins Viton oil control O rings.
Cheaper non inconel springs should not go in any 13BREW motor.
Once soda blasted the rotors are weighed on my gram scale. I don’t look at the factory letters as I have found they don’t necessarily reflect actual weight.
The gram scale is your friend.
The next step is to final build the front end which includes the oil pump. Oil pumps need to be checked for clearance.
Rotor housings are both the most expensive (almost $900 each) and most often degraded single item. Lubricating the apex seal contact area of the housing has been a challenge that has only been partially accomplished.
The 13BREW has one lubricating hole on the rotor housing. Later Renesis motors started with two holes and added a third. Of course the better solution is to premix, something Mazda couldn’t ask non gearhead owners to do… and even premix doesn’t get it done 100%. Far from it. Most incoming rotor housings look chromey and have chatter marks. While chrome might look good on bumpers (back in the day) it provides clear evidence of degradation.
New rotor housings look like finish-honed cylinder bores on a piston engine. The color is grey with oil retaining hatch marks.
Chatter is another significant enemy. The apex seals are spring loaded and as such are free to move in and out. What would cause them to do that?
Spark Plug Mountain
You can lose 20/30% compression from collapsed corner seal springs.
I had been running the generic Atkins chrome looking corner seal springs which i believe are from the RX8. As a general rule stay away from any item from the RX8 engine. I subsequently learned that the OE FD springs are made from inconel, are a bit stiffer and are much more resistant to temperature. Spend the extra ten bucks.
Another item I am now using is the OE corner seal. While perhaps not as strong as the Atkins offering I feel the hard chrome coating offers less drag with the side iron and corner seal well.
Here’s a picture of the corner seal and spring. Note the color of the inconel spring…
Add a new pilot bearing and pilot bearing seal, all fixed.
Recently the rear oil seal has been in the news. Atkins a few years ago had switched to using the RX8 seal which is half of the seal depth of the Orange FD seal. Do not use the RX8 seal as it has a tendency to leak. DO NOT use a flat blade screwdriver to remove the front and rear seal as you WILL score the mating surface and you will create an oil leak.
Next up, straighten the oilpan which is always bent during removal.. Plan B is new oilpan, not a bad idea as they are inexpensive.
I always remove the drive hardware for the evil external oil pump and replace the oil pump pressure spring as they are generally short by .2 of an inch.
Having bolted the block together it is time for a hold your breath moment: checking coolant system integrity.
Cap the heater pipe in the rear iron, install the coolant temp sensor in the rear iron, check the coolant drain bolt in the mid iron is closed and bolt the coolant pressure tester on the front iron. I have already plugged (drilled, tapped and an Allen plug) the throttle body coolant return line tube on the top of the rear iron.
I pressure the coolant system to 30 psi to check for leaks.
After swapping in new max clearance bearings we prepare the front stationary gear. I use two assembly lubes. The red lube is a WalMart grease from Shell. I use Driven assembly grease for bearings. Check out the Driven site. Driven was developed by Joe Gibbs Racing.
I use a set of water pump studs as mounting guides.
I removed the brace, checked to see if the pan rail lip was bent inward (often happens). Nope. The brace needed a bit of a tune up in that area. This would have been a great opportunity to strip the bolt.
Here’s the almost finished product. After installing the oilpan and brace i add the motor mounts to aid overall silicone setup. The inner silicone takes a week to fully set.
We are now close to buttoning it up by clamping 5 plates together that will make between two and three times the power of the factory motor.
We are faced with a familiar question:
Will the Mazda component, in this case tension bolts, work properly at elevated power levels. I see evidence on virtually every incoming motor that there is a need for increased clamp.
We do have options. Numerous kits of 16 studs are offered that require no machine work and are 10 mm rather than the OE 9.5. Others offer a smaller number of bolts that increase clamp in the lower section of the motor where heat and compression rule. There are other options that add additional tool steel dowel pins.
SInce my power neighborhood ends around 600 i rule out additional dowel pins and any item that requires machine work to the stack. Perhaps fine if you are looking to make monster power. I will also pass on anything that doesn't deliver equal clamp around the entire periphery.. in other words i want 16 10 mm tension bolts.
On the suggestion of my 1000+ hp friend i swapped in a set of titanium tension bolts and ran them for a few years. I wasn't happy with the threads and replaced them with another set of aftermarket chrome moly studs.
I retained an uncomfortable feeling re studs and the OE tension bolts... they seemed sort of like a black hole. Kind of like oil. Un-ending amounts of opinion but lacking a data driven underpinning.
Unlike oil where there are no easy answers, tension bolts/studs are easily rationalized. I elected to get the answers. I sent multiple OE tension bolts and 3 aftermarket stud offerings to a certified lab to test Yield (the point of stretch where elasticity is lost) and Tensile (actual breaking point) Strength.
The testing results delivered surprises and provided a clear path forward with re to bolting our higher powered motors back together.
Turblown's studs delivered as advertised and were in another zipcode versus others tested as to Yield ( over 185,000) and Tensile ( over 190,000) strength.
All rotor housings need to be measured as to width. The first measurement on my 6 page spec sheet is housing width in 4 locations. As per the Factory Service Manual, they may be no more than .0024 out of parallel. Out of parallel housings tend to lose outer coolant rings when making significant power.
I found one of the new housings for this particular motor was .002! My vendor couldn’t believe I measured new housings and also was shocked that they were almost out of spec. He sent me another housing and asked me to send the housing to another shop that had ordered new housings… hmmm
Here’s a picture of an incoming rotor housing being measured for width.
Properly prepped (shiny) oilpan rails
I continue to like Goopy apex seals. I have a very sophisticated/accurate knock system onboard (see Tuning/ Hardware Section) and have gone thru a number of detonation incidents over the years never breaking or warping one of their seals. i also like and use Rotary Engine, E&J and RXParts.
Apex seal grooves over time widen out near the corner seal area. Almost every motor I see has .006 apex seal to groove clearance at the outer 25% of both grooves. Generally 3 to 4 thou in the middle 50%.
Most builders ignore this issue. Goopy solves the problem with an optional .004 oversized seal. Brilliant! The builder is required to resize the apex seal grooves and the corner seal grooves. It takes me a couple of hours but the end result is .002 seal to groove clearance the length of the groove.
After a day’s labor the rotors are ready to go:
Issues ALWAYS surface on EVERY build that are not in the book and they must be recognized and defeated.
Note that the oil pan lip is not allowing the brace to bottom against the rail. The last thing you would want to do is try to (over) torque it down as the bolt would strip the aluminum threads in a heartbeat.
Notice no carbon on the faces or inboard of the sideseals. Tight sideseal clearance gets it done. The motor gained compression during the four years.
Front iron, front rotor and front housing ready for the crank and intermediate iron:
This is why you weigh. Here are a pair of rotors (an “A” and a “B”) from an incoming motor. The letter designations indicate they will be fine. The scale says otherwise as they are 48 grams apart! No wonder the rotor bearings showed copper.
Rotor bearings are often in good shape. If I see no copper or abnormality and they produce proper clearance they need not be changed out.
Rotor bearings have a 72% larger diameter than main bearings and rotate at one third the speed and therefore lead a fairly easy life.
The steel oil control rings are similar to the lower oil control piston ring in that they primarily control crankcase pressure/blowby. It is important that they be on spec with regard to wear or you will be filling your oil catch tank in short order. Oil control rings are fairly expensive and unfortunately it is important that they be on spec. Much of the time I find them O K to reuse.
Notice the lateral carbon deposits flanking the lead sparkplug boss. Clearly the apex seal is NOT contacting the surface at that point. The reason is that the area immediately around the plug is overly hot, needing to expand and it expands upward like a mountain. This forces the apex seal off the housing losing compression and starting continuous oscillation which eventually degrades both the apex seal and the housing. This is one of many important reasons to run 10 or colder heat range. Colder plugs reduce the height of spark plug mountain.
Apex seal oscillation from Spark Plug Mountain causes dreaded chatter marks which destroy compression.
The twin evils are Chrome and Chatter
Type your paragraph here.
After soda blast cleaning, porting and (optiona)l paint we start with the front iron. The OE coolant seals and compression rings are installed using Hylomar.
While the crank is pretty much indestructible I do find around 20% fail the runout spec.
It may be as simple as the too much drive belt tension as almost every front main bearing shows wear/copper at 11:30 looking at the motor from the front. This is caused by over-tightening the water pumpdrive belt.
The win win solution is the addition of a Full Function Engineering water pump idler pulley. (More soon in the Tuning/ Hardware Section).
BTW, copper at the front at 11:30 is mirrored by copper at the rear at 5:30.
(Why do Newton’s Laws always pop up?)
Back to our stack: