Question about re torqueing heads

Corky · 11-19-2006, 06:52 AM

Can any of you guys give me advice on retorquing the heads on a 2004 R1100s? I had the rings replaced and will need to retorque the heads soon.. I prefer to work on the bike myself. I know its not a big deal but I have never used a Torque angle gauge. I've retorqued a few heads in my life, just never with this gauge. Any advice on the procedure would be apreciated.

Thanks,

Corky

Moybin · 11-19-2006, 08:07 AM

Corky:

All torque wrenches require using your other hand to apply counter-torque to the head of the wrench, otherwise the value will be incorrect. To do this you push away from yourself against the head of the wrench while pulling the handle toward yourself. Among other things this counteracts the tendancy of the bolt shaft to twist toward you.

The value you want is a "running torque" value. For instance, if you need 10 ft-lbs of applied torque, the indicator will "hit" 10 while the wrench is still moving.

"Clicker" torque wrenches do this naturally, gage wrenches take a bit of practice.

If you are using a gage torque wrench that has a + and - range scale AND an indicator needle that you can preset, then the best way to use this wrench is to set the torque value you want with the needle, say 60 ft-lbs. Then apply torque until the gage hand reaches the needle. This way you don't have to sit there trying to remember what torque you are supposed to be achieving.

Bob Hancock · 11-19-2006, 08:58 AM

I think he's talking about the angle torque business needed to retorque the heads. I haven't done this but it seems easy enough. The book says:
" 1. Tighten the cylinder head nuts one after the other in a diagonally opposite sequence.
1.1 Slacken off one nut at a time.
1.2 Tighten nut to initial torque of 20Nm.
1.3 Tighten nut to specified wrench angle of 180 degrees."

This tells me that after the loosening and tightening to 20Nm, then just set your breaker bar to 9 o'clock and turn it to 3 o'clock. I have the tool to do it by degrees, but a wrench friend told me the 9 to 3 breaker bar would be fine. Good advice on holding the wrench (or bar) though.
Hope this helps.

Moybin · 11-19-2006, 09:02 AM

Yup, Bob, forgot about this "new" method. It uses the plastic qualities of the bolt to achieve the proper compression. One has to wonder what the "real" torque happens to be when done.

ps: if the manufacturer has a specific torque pattern (I'm sure they do) it must be adhered to or "bad things" can happen.

roger albert · 11-19-2006, 09:20 AM

Bob is just right.

As for the real torque, you never know it with the old method either. In fact, you 'know' it even less with the plain wrench method, which is why angle torque is better.

Here is why.

1st, you have to know that we're using friction (when the wrench clicks or how much the needle bends) as a proxy for torque, which is itelf a proxy for tension in the bolt, which is used to achieve clamping force (the ultimate goal) The whole thing is, to say the least, very indirect, and has, when you throw in wrench design, calibration, tolerance, and fastener surface, finish, and lubrication, a VERY wide range of variance. 20-100% errors are not common.

Anyway, the errors in implied torque (from friction) are quite a bit lower at low clamping forces, than at high ones. The amount of additional tension achieved on a bolt for a fixed amount of additional fastener rotation, on the other hand, is very repeatable and predictable, especially after some initial preload had been applied. The modern angle torquing method takes advantage of both of those. You get "in the ballpark" of a known "1st torque/tension value" fairly accurately at the low initial adjustment (remember, the inaccuracies are less at low values) So now, you have a pretty good idea of where you're at. A pretty darn precise and predictable additional torque is then applied by moving through a fixed arc, thus giving a better defined and controlled final clamping force.

The method uses a torque wrench in traditional application where it is more accurate (low friction / tension) and the angle method where it is most accurate (incremental tension beyond initial preload)

Neither method is perfect, but the angle method is more or less teh best of both worlds.

hope that helps. In any event, Bob's functional description is spot on. It looks like and extra step and more work, but there's really nothing to it. Just torque like you used to (using Moybin's good description of conventional torque wrench application) but to a lower value. The turn it some more. Usually it's spec'd at 90 degrees. Even w/o a gauge, the human eye is pretty good at noting 90degrees (1/4 turn) You do it a lot when fiddling with carbs.

R111S · R111S's Garage

Thanks for the explanation roger...I've learned something good to know today thanks to you.

I've worked around Mechanical Engineers for 20 years and was even helped to do a "bolted joint torque analysis" for a Fuel Filter Differential Pressure Sensor mounting block I designed for the Caterpillar 3500 Diesel Engine...the first and last piece of "Iron" I've ever designed...I'm a "EE" so I don't get the chance to do that very often if at all. I currently test Software on Embedded Controllers for Back Hoe Loaders and Skid Steers at our proving grounds.

No one ever explained about this method of torquing a bolt...how long ago did this particular method come into practice?

Anton · 11-19-2006, 12:22 PM

Quote:

Originally posted by R111S

No one ever explained about this method of torquing a bolt...how long ago did this particular method come into practice?

It's been around for a while, at least for as long as I've been playing with engines. Usually used only for important bolts, head bolts, flywheel bolts, etc. due to the extra step. I'm assuming Caterpillar uses it, I know your competition does. For head bolts it is often used in conjunction with torque-to-yield bolts, which are plastically deformed for even better control of clamping force.

If you are in the big engine business and want accurate control of clamping forces, you switch to hydraulically stretched studs.

http://www.freepatentsonline.com/6167764.html

Corky · 11-19-2006, 03:10 PM

Thanx for the replies... Its the angle part that was new to me. I uderstand that now. But I'm wondering about the loosening of each bolt before you bring it to the proper torque. You do each bolt completely (loosening, torqueing and the extra degree cinch) one at a time?
Meaning that after you finish torqueing and the extra degree tightening of one specific bolt, you then loosen the next bolt in the sequense and go through the same procedure? Just seems a bit odd you would have a bolt brought to the proper torque and the one next to it would be so loose..... Anyway I don't want to make anything probably so simple, difficult. I have an unabused Snap On clicker wrench. Think I'll buy one of the degree gauges and have a go.
One thing I like about a bike with Jugs. Sure make this kind of work easy...

Thanks for the help!!!

Much appreciated,

Corky

PS> Riding in the CA Sierra Nevadas was awesome today. With the late winter (no winter) we are having, Montor Pass was epoch today. Totally deserted and perfect road conditions!! ZOOM!!

Moybin · 11-19-2006, 04:35 PM

Quote:

Originally posted by Anton
... If you are in the big engine business and want accurate control of clamping forces, you switch to hydraulically stretched studs.

Commercial nuclear reactor vessel heads are held together with studs about a foot in diameter that are hollow. They have these HUGE electrical heating elements that go into them and heat them to something like 400 F, then the nuts are put on to about 98,000 ft-lbs (yes 98K!!!) with hydraulic drivers and torque multipliers, then the heaters are removed and the studs SHRINK, leaving the "torque" at about 1,000,000 ft-lbs EACH!

Needless to say each nut and stud has a serial number and is matched for life. They even do dye penetrant and magnetic testing for cracks.

Anton · 11-19-2006, 05:03 PM

I suppose by "big engine" I was thinking 150 to 500mm bore. Moybin's talking about FREAKING HUGE applications. For those you need to be more creative.

Bob Hancock · 11-19-2006, 05:32 PM

Kinda reminds me of the pics of those enormous ship engines we've seen here in the last year or two....the ones with people climbing out of the sump. Betcha one of you guys still has that pic. Wonder what the torque specs for those main bearing rods are?

Bob Hancock · 11-19-2006, 05:36 PM

Found it!

R111S · R111S's Garage

Quote:

Originally posted by Corky
You do each bolt completely (loosening, torquing and the extra degree cinch) one at a time?

I don't think this is correct...shouldn't each step be performed on all bolts (using "star" pattern) before going to the next step in the torquing process?

Yes a big motor indeed! I was at the Shipyards in Bremerhafen, Germany a few years ago. Caterpillar (my employer) owns MaK Motoren GmbH in Keil, Germany. They make 6, 7, 8 and 9 cylinders-in-line, and 12 and 16 cylinder-Vee-type engines up to 16000kW (yes that's 21456 HP!).

We were there looking over the wiring of four MaK Diesel Propulsion Engines on a ship that was still under construction...I was awe struck to say the least. We walked up 10 flights of steps going from the engine bay to the upper deck...what a view of the harbor. Seafood was awesome too!

http://www.mak-global.com/home/home.php

Bob Hancock · 11-19-2006, 06:03 PM

The way I read the instruction is that you:
1.1 Slacken off one nut at a time.
1.2 Tighten nut to initial torque of 20Nm.
1.3 Tighten nut to specified wrench angle of 180 degrees."

The instruction about the diagonal pattern is a general instruction. The sequence doesn't begin until 1.1.

Well at least that's how I would do it....but then what do I know?

roger albert · 11-19-2006, 06:39 PM

Bob is Right

zx9rmal · 11-20-2006, 08:14 AM

Quote:

Originally posted by Bob Hancock
Found it!

WOW!!! The new Harley crank!!

zx9rmal · 11-20-2006, 08:17 AM

And, with my R12S, I need to get that dumb BMW tool to remove the coils on the plugs. Is there an alternate method of removing the coils?

anotheroldrider · 11-20-2006, 08:37 AM

As the parts and gaskets settle in together and the bolts take up the initial slack in the threads, etc, getting everything somewhat evenly compressed is an important starting point. I ease up to the starting point (20Nm in this case) in a couple steps around that sequential pattern with a click type wrench just to make sure things are uniform before starting steps 1.1 - 1.3.

I also prefer the "torque angle" or "turn-of-the-nut" method, for exactly the reasons Roger described. Even very minor differences in thread quality, lubrication, etc can trick a torque wrench - all it measures is the amount of friction it takes to turn the fastener. Turn of the nut methods (or as BMW calls it, Torque Angle) are far more accurate because they're calculated on the exact thread pitch to determine tension - 1/2 turn = a specific amount of bolt stretch (as long as you know how tight it was to begin with).

On the subject of torque wrenches, get yours calibrated once in a while, and take good care of it. I have the occasional good fortune to have access to calibration labs at work, and it's very common to see wrenches come in with 15-25% error. Also, avoid using the wrench at the extreme ends of the range - I have three wrenches to get all the torque values I use frequently.

Phil Buttolph · 11-20-2006, 10:53 AM

Fascinating thread..makes sense...calibration of torque wrenches has always seemed a weak point in assemble..

Question: For the metric impaired, "20 Nm"? Ft/lbs..Inch/lbs value? and I don't clearly see how once you are at 20 Nms on all bolts using admittedly an inaccurate method [torque wrench]..that a 180 degree tightening of the bolts will give you an accurate bolt stretch/torque point? Seems as though the torque wrench error would be the LCD.

Perhaps also answered..but do you cross pattern going in and going out and then finish with the final 180 degree tightening? Thanks, we have a lot of collective wisdom and I appreciate that..

aaronsb · 11-20-2006, 01:58 PM

I loosen and retorque one bolt at a time, in the pattern designated by the factory. If it's a single cylinder like a beemer, cross pattern. If it's an inline somethingerother, I'll start from the middle and work out. But I only loosen and retorque one bolt at a time. If you loosen all the bolts at once and then put them back, it's not going to do much good for "retorquing" the head now, is it?

Quote:

Originally posted by Phil Buttolph
Fascinating thread..makes sense...calibration of torque wrenches has always seemed a weak point in assemble..

Question: For the metric impaired, "20 Nm"? Ft/lbs..Inch/lbs value? and I don't clearly see how once you are at 20 Nms on all bolts using admittedly an inaccurate method [torque wrench]..that a 180 degree tightening of the bolts will give you an accurate bolt stretch/torque point? Seems as though the torque wrench error would be the LCD.

Perhaps also answered..but do you cross pattern going in and going out and then finish with the final 180 degree tightening? Thanks, we have a lot of collective wisdom and I appreciate that..