Mazdaspeed Forums

Mazdaspeed Forums (http://www.mazdaspeedforums.org/forum/)
-   MazdaSpeed 3 - Engine, Transmission & Driveline (http://www.mazdaspeedforums.org/forum/f10/)
-   -   Cylinder head machining limit (http://www.mazdaspeedforums.org/forum/f10/cylinder-head-machining-limit-154775/)

99pgpgt 09-09-2013 05:55 PM

Cylinder head machining limit
 
Hey guys, I joined here for a friend who owns an 07' speed3. The car was purchased a few months ago with some engine work previously done. The engine has powdered rods and wiseco forged pistons. The head was slightly off and needed machined. Apparently, the machine shop took off .010"...That seems a bit much.

Now, my buddy had problems with the pcv lines being routed very oddly. And the car started acting up. (Unfortunately I have no pictures of how the pcv lines were routed due to the fact that he didn't alert me of a problem until he started tearing the thing apart) So, he has the turbo rebuilt by a local shop (xsboost) because he thought it was bad at first. Put the turbo back on and still didn't feel right. Ended up getting oil into the cylinder he believes due to the pcv routing and blew a headgasket.

He took the head off and there was coolant in the cylinder 2 and 3. It was an mls gasket used before. So now it comes down to him taking the head into a machine shop. If the head does need a resurface, what is the limit on how much more can be taken off. And, would it be a good idea to use a 5 layer mls instead of a 3 layer?

Thanks guys.

phate 09-09-2013 10:48 PM

I don't think anyone knows exactly how much would be ok to take off - we don't know how thick the head deck is. I had .008" taken off of a head (works great), and my machinist said it would probably be ok taking upwards of .010" off of it. If it's not going on an engine that will see severe duty, you might get away with taking more off of it. Heck, you might be ok, regardless.

If it was me and I didn't have the cash to source a different head, and I knew it wasn't going to get thrashed all the time, I'd be ok taking a few more thousandths off of it. Before you get it milled again, make sure the cams don't bind from warpage.

I'm a fan of the OE gasket. If you want to keep the compression closer to stock, then maybe consider the thicker gasket. The bump in compression from milling .010" was something around .3, so you may end up just under 10:1. Not a bad thing, especially if you have decent fuel available.

99pgpgt 09-09-2013 11:55 PM

Thanks for the input.

I wish he would have waited for me before he tore it down. He took out some of the valve buckets and didn't number anything.. So putting it together will be fun. Thankfully I found the thread on here of another member assembling a 2.3 from the bottom up. Along with how to time it which I'm thankful for.

I am hoping the head is still straight and he can get away with only taking a thousandth or two off. I'm not certain on that 10 thousandth resurfacing done before as I haven't spoken with the cars previous owner, or the shop that did the engine work.

hnda etr 09-10-2013 12:55 AM

Edit: I'm dumb.

Tokay444 09-10-2013 04:09 AM

The buckets are sized in decimal millimeters, but they aren't numbered per cylinder.

silvapain 09-10-2013 06:49 AM

Even if you mark the buckets and put them back in the same spots you still have to measure clearance between the buckets and the cam lobes. They WILL be off.

99pgpgt 09-10-2013 07:27 AM

Quote:

Originally Posted by silvapain (Post 2248645)
Even if you mark the buckets and put them back in the same spots you still have to measure clearance between the buckets and the cam lobes. They WILL be off.

Ok, that's good to know. He works full time so getting time to work on this thing is few and far between.


All times are GMT -7. The time now is 07:50 AM.

Powered by vBulletin® Version 3.8.6
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.
SEO by vBSEO ©2011, Crawlability, Inc.
vB.Sponsors

©Copyright 2008 ; 2019 Cymru Internet Services LLC | FYHN™ Autosports HQ
Ad Management plugin by RedTyger

Page generated in 0.17455 seconds with 11 queries