Is there a DIY way to check a head for cracks?

E28 technical advice asked and given! Troubleshooting, modifications and more.
Post Reply
turbodan
Posts: 9701
Joined: Jan 09, 2007

Is there a DIY way to check a head for cracks?

Post by turbodan »

Other than a visual inspection, is there any way to check a head for cracks without x rays or anything crazy like that?
Mike W.
Posts: 27986
Joined: Feb 12, 2006
Location: California Whine Country

Post by Mike W. »

I can only speak about M30's, not M20's, which is what your question is probably about. But... On M30's, with the design of the cooling system, bubbles in the coolant tank indicate a cracked head or blown head gasket. Not turbulance, and it can be hard to tell the difference, but bubbles. They obviously come from the combustion gasses being pushed from the combustion chamber to the water jacket and have to escape somewhere.

You might be able to do the same thing with compressed air in a spark plug hole also.

M30's have such a familiar pattern of cracking, and I've seen so many, it's usually pretty obvious to me. They always crack between the exhaust valve seat and the water jacket. Later B35's I'm told will crack underneath the cam, but I've less experience with them.

The way machine shops check is just by pressure testing with air in a water bath and looking for bubbles. There are more exotic dye and powder kits, but I've never used them and I don't know how well they do work. But visual's always worked for me.
M. Holtmeier
Posts: 3055
Joined: Mar 11, 2007
Location: Josephine, Texas

Post by M. Holtmeier »

You'll have to make a metal plate to cover the mating surface of the head. Cut out the chambers to expose the valves, drill holes for the head bolts. Block everything else that holds coolant. Find a piece of rubber to seal the plate against the head and bolt it up. Making a plate to block off the t-stat housing will allow you to add a fitting pressure check the head. Get an air regulator, fill the head up with water and pressurize. You might have to roll the cam over to inspect all the runners. Sometimes you have to bring the head up to temp to get them to leak also. Things like this are a bitch to make the first time, but when someone else needs it, you drink for free.
superbeeblaze
Posts: 242
Joined: Sep 02, 2007
Location: KANSAS

Post by superbeeblaze »

is this on or off the car?
Shawn D.
Beamter
Beamter
Posts: 22375
Joined: Feb 12, 2006
Location: Alpharetta, GA
Contact:

Post by Shawn D. »

Mike W. wrote:On M30's, with the design of the cooling system, bubbles in the coolant tank indicate a cracked head or blown head gasket. Not turbulance, and it can be hard to tell the difference, but bubbles. They obviously come from the combustion gasses being pushed from the combustion chamber to the water jacket and have to escape somewhere.
Yeah, I thought the overheating and bubbles in the coolant tank on my 535i were due to a blown head gasket. However, the real culprit was a clogged radiator.
elwessoE28
Posts: 249
Joined: Oct 14, 2007
Location: Indiana

Post by elwessoE28 »

If its on the car, couldnt you just use a pressure checker to see if it holds pressure?
turbodan
Posts: 9701
Joined: Jan 09, 2007

Post by turbodan »

The 885 M20 heads are also supposed to crack under the cam between the journals around cylinder #4. I checked there, but its not always easy to spot a short hairline crack across a sand cast finish, at least until it cracks across one of the cam journals. I think it should be more apparent now that it seems to have gotten much worse.

It sounds like a huge PITA to make sealing plates to actually pressure test the head. Hopefully its obvious upon disassembly. If I cant find anything it might come to that though. Its all still in the car right now. Looking at a teardown this weekend.
slave2school
Posts: 371
Joined: Aug 23, 2007
Location: NS, Canada

Post by slave2school »

I wonder if some company makes a liquid penetrant NDT type kit you can use with a standard black light.

I know there are UV liquids for rad leaks, maybe something like that can be used to help identify any trouble spots.
Bimmerguy2002
Posts: 11973
Joined: Oct 03, 2007
Location: Lodi, California

Post by Bimmerguy2002 »

what sucks is that you cant magnaflux a head since its aluminum but there are other ways on the cheap that machine shops use.
Making a fixture is the most effective way as someone else mentioned, bolting the plate to seal the head so the fluid cannot escape through the jackets on the bottom of the head, thats time and money for a shop to make it.
You can use penetrating dye for aluminum heads under a UV light, you can even use a black light for certain dye's.
turbodan
Posts: 9701
Joined: Jan 09, 2007

Post by turbodan »

I actually did end up finding a crack in ths head. When I got the valve cover off I found a little puddle of blue coolant sitting right over cylinder #5. So I figured that gave me a good idea of where to look. It hadnt cracked across the cam bearings, unlike the other junk 885 head I've got, so it was incredibly hard to see. I started at it for probably 15 minutes before I spotted it.

Just like the other cracked head I've got, the crack runs along the top of the water jacket, pretty much following the peak in the casting under the cam, right below where the intake valve springs sit on the head. It goes all the way across the top of #5. I'd take a picture of it, but it would be impossible to see.

If the car ran longer it would have built up a line of gunk along the crack where the coolant was mixing with the oil. This happened with the other junk head I have. If you run it that long you'll probably destroy the bottom end bearing shells too, so hopefully you dont see a little brown line along the crack.

I thought it was a head gasket though, and I was wrong. So if your M20 turns up with water in the oil, take a very very close look at the head. I know its not common, but it does happen, and it does suck to put it all back together for nothing.
BudFox
Posts: 655
Joined: Feb 07, 2007
Location: Chicago, IL

Post by BudFox »

Just curious...how many miles on the cracked head?
turbodan
Posts: 9701
Joined: Jan 09, 2007

Post by turbodan »

This one was at about 250,000. But I'm sure it wasn't due to age and fatigue as much as it was blowing the head gasket running up to 20 psi. On a normally aspirated car it would have probably been fine pretty much indefinitely.

Usually the only way to crack a 2.5i head is to severely overheat the engine. I believe my problem was that it was blowing combustion gases into the coolant, which was displacing the coolant flowing around #5, causing the localized overheating and the crack.

So dont run a car on a blown head gasket and you'll probably be alright. I dont know what I was waiting for the replace it, but waiting until there was coolant in the oil was a bad call. Its hard to get that kind of mixing at the head gasket on an M20. It will leak externally well before it mixes coolant in the oil.
Patrick McHugh
Posts: 4163
Joined: Feb 12, 2006
Location: Richmond, VA

Post by Patrick McHugh »

Here's what we use to test some aviation parts for cracks. Should work fine on a smooth surface.

http://en.wikipedia.org/wiki/Dye_penetrant_inspection
turbodan
Posts: 9701
Joined: Jan 09, 2007

Post by turbodan »

Thats a cool looking product. It would be interesting to see how this crack stands out with this dye...
Patrick McHugh
Posts: 4163
Joined: Feb 12, 2006
Location: Richmond, VA

Post by Patrick McHugh »

Here's how it turns out usually.

Magnaflux is the brand to look for, they sell kits for a hundred dollars or so, it can be pretty handy!

Image
Post Reply