the Miata setup and preparation portal Flyin' Miata tech line 970 464 5600 

Everything you ever wanted to know about the 99 Miata thrust bearing failures

Some '99 and '00 Miata's are afflicted with the "mis-machined #4 main bearing cap" and there is a lot of mis-information and associated panic because of it. The Mazda technical service bulletin is listed at

The blocks we have measured that haven't totally eaten themselves (rare, as by the time we see them, the crank has usually machined its way right in to the block) measure to factory specs. This would point to the factory bulletin being a mis-nomer and the problem actually lies in defective or improperly sized thrust bearings or possibly just a less than stellar design. We have seen this problem on engines built long before '99, including 1.6 factory crate engines. As displacement and power have gone up, Mazda has used heavier clutches with the same half circle little thrust bearing until this issue prompted them to nearly double the surface area of the bearing sometime in the 2000MY. Many, if not most other engines use a full 360 degree thrust bearing that provides a lot more surface area. Mazda's post 2000 fix is still a 180 degree design, but it does provide much more surface area than the old style. Skimpy old style in top of picture:

<BC> What does a thrust bearing do, why do cars have them and how can we keep them alive?

Something has to take the end forces on the crankshaft. What causes these end forces? A small amount can be induced by pulley mis-alignment, although not likely with only factory belts and accessories. Not all aftermarket superchargers are designed to factory levels of quality, so belt mis-alignment is inherent in some kits. This is not the main culprit in my opinion.

When the safety nannies decided it was too much to ask us to make sure we were in neutral (or had the clutch down) before starting the car, the manufacturers started installing clutch interlock switches. A clutch interlock requires you to push the clutch pedal all the way to the floor before the starter will work. Sounds like a reasonable idea-probably helps keep Darwin at bay, BUT....When you step on the clutch, the effort of your foot is multiplied many times by the ratio of the clutch pedal itself, the clutch hydraulics and, down in the transmission, the release arm. Putting fifty? lbs of effort into your clutch with your foot may put close to 2,000 lbs of pressure on the thrust bearing. This is bad enough each time you use the clutch when the engine is running, but imagine what happens when you start the car.

Here are your unsuspecting crankshaft and thrust bearing, perhaps napping for the 10.5 month Winter in Northern Canada. Whatever oil was between the crank and bearing when the engine was last shut off have probably drained long ago (remember, the crankshaft sits above the level of the oil in the pan). Now you come along with your big foot (or, through the miracles of leverage, even your dainty foot) and squeeze whatever last bit of oil was still in there out with perhaps 1500-2000 lbs of force. Now you start the car and the poor, unloved and unlubricated thrust bearing gets ground on by the crankshaft until the oil pressure comes up. How would *you* feel?

Heavy duty clutches can exacerbate any thrust bearing problems. Our advice at FM is to assume a little responsibility for your own actions and either jump the wires on the starter interlock switch or use mechanical means to hold it down. This means your car will start without your foot on the clutch. People managed to do this for many, many years before interlocks came along, are you any dumber? All FM performance clutch kits come with a clip to accomplish this task.

Mazda's solution to this is to warrantee a whole new engine if there is a problem with an individual engine. Not all engines are affected. Many people, however, have gone into dealerships demanding a new engine because their VIN number fell into the effected range when, in fact, their engines were just fine. <As I update this information in late 2009, it's probably safe to say Mazda won't be buying you anything for your 1999 Miata.< BC>

Thrust bearings. click for larger image
 thrustbearings2.jpg (38 KB)  thrustbearings3.jpg (38 KB)  thrustbearings4 (38 KB)

The engines are destroyed by a self-machining process which adds metal shavings and metal powder to your oil. Thrust bearings come in 4 thicknesses from the factory; 2.5mm, 2.625mm, 2.75mm and 2.875mm. Thrust bearings that were too thin may have been installed in a batch of engines regardless of what was required (that may be why not all engines are affected, some needed the thinner bearings anyway). For those engines that required a thicker thrust bearing installed the greater tolerance allowed the crankshaft to move backwards and forwards to the point where the thrust bearing itself starts to have accelerated wear. (Most automatic equipped engines will not suffer any ill effects of a too small thrust bearing because automatic transmissions don't put any fore and aft load on the crankshaft like stick shifts do).

 thrustbearings.jpg (38 KB) Thrust for larger image

Once the thrust bearing has worn to a sufficient thinness the crankshaft will actually come in contact with the block and the main bearing caps and will start to self-machine them. The crankshaft is a far harder steel material than the cast iron block and caps so they lose this battle.

capcompare.jpg (27 KB) Compare normal main bearing caps with ones machined by the crank. click for larger image

blockgrooves. (29 KB) The block also gets grooves cut in it usually making it unusable. click for larger image

You can tell if you have an effected engine by checking the crankshaft endplay. This can easily done by *lightly* (it doesn't take much force) prying on the crankshaft pulley and seeing if it moves fore and aft.

The affected 99-00 engines were being delivered with .010 - .020" endplay tolerance from the factory. It should be no greater than .006" new. It is my recommendation that anything more than about .008" - .010" cold on a fairly *new* engine and you should be concerned. Mazda recommends .3mm (.0118"). If your endplay is less than .008" cold then forget about it and just drive the car.

After the thrust bearings have been worn enough they will be actually thin enough to fall out of their landing. Sometimes you get lucky and they fall directly into the oil pan, other times they bounce around for a while and mangle most everything, which means, even with a rebuild, you will likely have to get new pistons and rods too. But that has little meaning anyway since, without thrust bearings, the crankshaft can move 1/4" (which is why people report less clutch travel as a symptom) and nothing will hold it back from machining all in its path. In extreme cases rods have even broken because they are now out of alighnment with the cylinder bore.

thrustbearingclearance (31 KB) Without a thrust bearing the crankshaft can move up to 1/4".

Question: My 99 engine has XX,XXX miles on it. I just checked the endplay and it is fine now but do I have to worry about it in the future.

Answer: Really no way to say. If endplay is currently good, you are *probably* ok, BUT- it is not a great design to begin with. See notes on starter interlock above.

Question: Is there a way to mic / measure the #4 main bearing support machined recess area (where thrust bearings are fitted) and the thrust bearing thickness to determine the correct / appropriate guage bearing required?

Answer: Not without dropping the oil pan, which requires lowering the subframe. Just use the endplay measurement. Provides the same results.

Question: Is there an aftermarket 'saddle' type bearing retro-fit available?

Answer: No. The issue has nothing really to do with the design of the thrust bearing being a shim style or a saddle style. While it is true a saddle style would not allow the thrust bearing to fall out when it got too thin, but it would still wear at the same accelerated rate if the endplay was too wide. A better solution would be a 360 degree bearing, but it would require machining the bearing cap.

Question: How reliable is the end play measurment in determining durability?

Answer: Very. It is the sole determining factor other than oil change intervals. If your endplay is fine, the then motor is in spec.

Question: Is the end play measurment a fool proof way to verify the correct thrust bearing selection?

Answer: Yes. Change your oil regularly. That and defeating the starter interlock are the best ways to keep the thrust bearing clearance from growing.

Question: My car is past warrantee and my crankshaft endplay is probably 1/16" - 1/8". Is the engine repairable?

Answer: Maybe...maybe not. Depends on how quick you caught it.

  • If, by some miracle, you caught it before there were any noises (and why would you?), perhaps. Keep in mind the oil goes to the other bearings and to the head. Once things start grinding, the cam journals will be damaged quickly. The journals are part of the head itself, so if they're bad, the head is bad. Short answer, pull some cam journals. If they look ok, pull the pan and have a look at the thrust bearing, bearing saddles and crankshaft.
  • Chances are slim that you have anything but a boat anchor.

In any case, you will need to remove the oil pan and inspect the thrust bearings, block, #4 main cap to make a determination.

miataoilsystem (38 KB) 1.6 and 1.8 'B' oil system. click for larger image

provided by Flyin' Miata | tech line 970 464 5600