Torquing down 4A-GE head bolts.

Hi all,

I posted this question over on the MR2 MK1 Forum too. The thread briefly came to life again recently which reminded me that I had written up what I actually did but forgot to post it. Well I found it and here it is...

-------------------------------

I've come to decision about torquing my cylinder head down. First of all, thank you all for your thoughts and information. No doubt whatever I've decided to do will upset somebody but that's life. The best I can do is explain my thoughts - right or wrong - and hope it's ok. And if it's not ok I'll admit my mistake and you'll all know what not to do.

To recap.

The 4A-GE manual written when my car was built says you should torque it to 43ft-lb over several passes.
After a couple of years Toyota redesigned the block slightly and the new manuals says you should torque to 22ft-lb, then add 90 degrees, then another 90 degrees. I don't know what torque this would actually give you but it'll be a big number. Massively more then 43ft-lb. I'd guess 75+ft-lb. Of course if the bolts start to stretch it'll limit the torque needed to tighten them.

The later manual does not differentiate between the two types of block. It just says 22ft-lb + 90 + 90. In all probability it's probably fine to use it on early engines too. Had the torque figures been 'reasonable' this would have been a no brainer.

The next debate is 'stretch bolts or not?' Most people say they aren't. A few say they are. At the end of the day if you tighten any bolt enough it'll deform. For me the most telling thing is that the manual does not say you have to replace them. That probably means they aren't being stretched even by the 22 + 90 + 90 method.

So where does this leave me? Torque figures from 43 to 75+? Mild possibility of stretched bolts. Plus different figures for other members of the Toyota A series engines or different manufacturers bolts. But all in the 45 to 75+ ft-lb range.

So I started going for the 22 + 90 + 90 method expecting to end up in the mid 40s. I torqued everything down to 22ft-lb and then added 90 degrees to the first bolt. The torque was up around 60 to do that and I got scared. So I backed it off to zero degrees and then went to 33 and 43ft-lb. Effectively the old school method. In new method terms this was about 22ft-lb + 80 degrees.

Now this is the spec it would have come out of the factory as. It kept the car running for 25 years and the gasket only failed when the cooling water escaped so it can't be that bad. Even though it's a little low compared with other head bolt torques of the day.

Then I stumbled into an old school mechanic who worked on these things at the time. He suggested 43ft-lb and then 'nip things down'. I'm not totally convinced by that plan either. But having pulled one bolt down to 22 + 90, squashed the gasket a bit, and backed it off again I wanted to 'make sure' it was all evenly tight. So after a few days sitting at 43ft-lb I set the wrench to 55ft-lb and tweaked it down a little more. That is 22ft-lb plus about 90 degrees.

So where does that leave me? Well tighter than the old method and not as tight as the new method. In fact I'm somewhere in the middle of range of values given in various places - but not actually on any of them.
But looking at the positive side, it's a bit tighter than the slightly loose old method, I haven't pulled the threads out of the block, the gasket hasn't been 'unclamped', and if it does leak oil into the water or visa versa I can always add the final 90 degrees to the bolts.

I've also been looking into the relationship between bolt torque and the joint clamping force it produces. Basically it's very random. There is a formula that allows you to calculate it but there is a 'fudge factor' to cope with friction under the bolt head and in the threads. And that's a massive unknown because it depends on how well the threads were cleaned and lubricated. Or even if they were cleaned at all.
Some people suggest you could be - I think it was - plus or minus 30% in error in the clamping force for a given torque. And that's with bolts installed at the same time with the same tool.
The problem is so bad at the extremes that it could lead you to vastly under tighten one bolt and break the one next to it. (This is bolts in general - not head bolts specifically.)
The 'snug it down and then add some degrees' method is more accurate (but still not brilliant) so that's likely to be why Toyota changed to it. What I haven't worked out it why they increased the torque (and presumably clamping force) so much. Or indeed if it pushes the bolts into their deformed state.

So there we are. I've done what nobody suggested and annoyed everybody! I like to be even handed.

Thanks again,

James

PS. I wrote that bit some time ago and forgot to post it. Since then the engine's had a few heat cycles and a road trip to the MOT station and back. Everything is fine. However, because the car failed the MOT it's not had a proper ragging yet. The jury is still out…