1997 Mazda MX-5 Miata: Assembly By Keegan Engineering, Pt. II
September 4, 2012
The role of proper hardware in an engine build is paramount, but the role of proper assembly is just as critical, and often understated. A few thousandths of an inch in the assembly of a high-performance engine separates a reliable, high performing engine from one that, well, isn't.
That's why we left the assembly of Project Miata's heart transplant, a 1.9-ish-liter long-rod VVT Mazda BP, to the pros at Keegan Engineering. We delved into the nuances of their work a bit in a recent short block setup entry.
Across the jump are a lot more insights provided by Keegan as the build process progressed. (I originally planned for this entry to get into the head, but I realized I had such a surplus of short block assembly notes that the head will have to wait for now!)
Mike Keegan torques the main bearing caps without the crankshaft in place in order to determine the proper bearing size to use based on his desired bearing clearance.
A note on the use of main bearing studs (not shown here but being used in this build): when replacing the main bearing cap bolts with studs -- even on a brand new engine -- you absolutely must first align-hone the mains. This is because studs provide higher clamping force (due to higher tightening torque) than do bolts, making the mains ever so slightly egg-shaped. Once align-honed with the studs secured, the mains again become round, and round bearings are happy bearings.
Mike recommends studs for BPs that will generate more than 250 hp.
A dial bore gauge is used to measure the installed bearing diameter. For this build the clearance will be set to 0.001" for every inch of crank journal diameter. Clearances are a balancing act -- going tighter increases load capacity at the expense of additional windage loss and reduced oil flow rate. Too loose and the bearing runs cooler and saps less power, but it loses load capacity.
We're using a stock Mazda BP crankshaft for this build. It's forged from the factory, and Keegan reckons it's a pretty stout piece good for many times the stock power output. Still, no dimension escapes scrutiny prior to use. Here Keegan measures the crank journals to determine their out-of-roundness, taper and surface finish. If any of those parameters don't meet his standards, he has the crank corrected.
ACL bearings are chosen for their high embeddability characteristics, high load capacity and size selection.
All cranks are rebalanced to Keegan's specifications. For inline-4s, he allows crankshaft unbalance of no more than 0.5 grams (at the radius). For pistons, the heaviest one can be no more than 1.0 gram heavier than the lightest. Same for rods.
Keegan is bearish about removing weight from crankshafts, particularly those of inline 4cyl engines, and double especially for those inline-4s that do not employ balance shafts. Like the Mazda BP, for instance. He'll do some bullnosing of the counterweights, but that's about it.
That brings us to three of the things Keegan views as common misnomers when it comes to high-performance engine assembly
1. Aggressively Lightened Crankshafts -- "Don't go too light," he cautions. "Inline fours are not as naturally balanced as boxers so they rely on more counterweighting." (He builds a lot of Subaru engines, too.)
2. Huge Ports -- "Bigger is not better." Read more about Project Miata's head work by Keegan here.
3. WPC Treatment On Everything -- "WPC is great. I use it on certain valvetrain parts to reduce friction. But people get carried away with it, WPCing everything, and it just isn't necessary.
Crankshaft thrust (aka axial) float measurement. The target for our iron block is <0.005" float, dry (an oil film would take up space and alter the measurement). Aluminum block engines generally require increased thrust float when cold as the gap shrinks more when the block heats up (aluminum has a higher coefficient of thermal expansion than iron).
As for thrust bearings, factory Mazda thrust bearings are the way to go here.
A fine stone deburrs the piston rings prior to use.
Then the bare rings are set into the finished bores. The ring gap is then measured and adjusted independently for each bore-ring pairing.
Piston pin bore measurement in action. For this build, Keegan will target 0.0005" wristpin-to-bore radial clearance (as measured cold).
Checking the pin bore diameter of the Mil.Spec connecting rods. These rods have pin bushings made from Ampco 45, which is a specialized nickel-aluminum bronze material used in heavily loaded applications.
Extreme applications that strive to remove every last bit of reciprocating mass can choose to run DLC coated wrist pins. DLC'd pins allow the rod's pin bushing to be deleted entirely, so the wrist pin rides directly on the steel surface of the rod. These pins are almost always made from titanium for further weight reduction.
In our case we're using 9310 case-hardened steel wrist pins from JE Pistons.
Note the cross hatching on the piston's pin bore. For those chasing every last bit of friction, Keegan says it is possible to use a mirror-finish on the pin bore (like that achieved by a diamond cutter) IF the block is equipped with piston oil squirters (ours is). If no squirters are present, mirror finish pin bores are a no-go, and he instead requires the pin bores to have crosshatching in order to retain an oil film during operation.
The wrist pin circlip should always be installed with the gap located at the "12 o'clock" or "6 o'clock" position -- if at 9:00 or 3:00, the forces acting on it during use could cause the gap to close up.
Use of a stretch gauge is important when installing connecting rod bolts. This is the optimum method to reliably establish the load on the fastener. Simply torquing them to a certain value is not as consistent as it is too reliant on the amount, type and application of lube on the threads. Stretch, however, is not. In addition, the stretch technique can determine whether a rod bolt has been yielded beyond its elastic range and is therefore no longer usable.
Our rods have ARP Custom Age 625 rod bolts, so ARP moly assembly lube is placed on the threads and under the head of the bolt prior to installation.
We're using a crankshaft damper (and 36-tooth crank trigger wheel) from BHJ Dynamics. Ours is an all-steel prototype; the production version is has an aluminum hub and steel inertia ring. We'll explore this critical piece more in a later post.
This is the coolant fitting that will feed the turbo. Early Miata blocks were pre-tapped for a turbo's oil and coolant. Our block is a later one that was not thusly machined/plugged from the factory, so Keegan machined the factory boss for coolant.
The stock oil boss (directly behind the red fitting), however, remains untapped. Keegan does not like how it would bleed oil pressure directly from the main oil galley at the rearmost main bearing. There's no sense depriving the heavily-loaded mains from any oil when there's a more benign source for the turbo's oil on the passenger side of the block -- at the VVT oil supply.
So, yeah, you could say Mike Keegan is thorough when it comes to his craft. The man is something of a perfectionist, and the care taken is the same for every engine that comes through his shop. There are no shortcuts.
Jason Kavanagh, Engineering Editor
Photos by Mark Takahashi
Keegan Engineering -- www.keeganengineering.com