: Uneven torque on main bolts?
Had a situation last night. When I timeserted the block, I only did 19 of the 20 holes because one looked like it had been plugged @ the factory and I was afraid after drilling there wouldn't be enough meat on the plug to hold the timesert. As I was running the bolts down the additional 65deg after the intial 15 ft/lb, the 20th hole (plugged) stripped. I had no choice but to try and insert, fortunately it worked. The problem I have is that when I was breaking the main bolts loose again, some of them seemed to be not nearly as tight as others, some required the usual amount of muscle to break loose where others were relatively easy....i'm stumped. I really don't want to put this back together till I figure out whats going on, any input here would be greatly appreciated.
11-25-04, 12:46 AM
You do seem to be having more than your fair share of trouble.
I'm really quite puzzled by the technique of a reference torque followed by given angle of rotation. I guess that it stems from the difficulty of getting a running torque, as apposed to static. I've asked before, but is there a set of torque figures issued in lieu of the angle technique? It would be a good check to see what the actual torque was as the angle was reached.
It seems to me that any variation in the thickness of the gasgate etc. in the initial stages would mean that the datum for everything after that was wrong. The relatively low forces at the start perhaps do not ensure a level playing field to begin from.
Clearly such a huge manufacturer would not be using the wrong techniques, so how does one check that the datum torque is accurate? Is it a running setting or static? Presumably no lube, but what about locking fluids that may affect the setting...are the threads always clean and dry when using Timeserts?
I would be interested to hear the history of this from bbob and others of why this technique is preferred over the old method.
Good luck RB
I think Bbob posted the reason somewhere a couple months ago, but can't find the thread. I also think my own stupidity is overwhelming sometimes.
I don't know what I was thinking, but when I put the case 1/2's together, I was torquing all the bolts down to 15 ft/lb THEN going the 65deg...The book does not call for torque/angle sequence like the head bolts, but I was treating it that way (somebody just slap me). It occured to me last night while pondering this that if I torque all the bolts and then rotate, that after the first couple of bolts the initial torque setting is not a constant....duh.
My mom in law had made a comment that she was really feeling sorry for all the trouble I was having but all I could reply with was..."Apparently God knows I'm doing this wrong and he's giving me signs to keep doing it till I get it right" The plus side to this is I've become very good at disassembling this and can strip a block in a matter of minutes.
At least with this attitude I haven't pushed the car into the street and set it ablaze....yet :lildevil:
11-26-04, 02:41 AM
The Supra twin turbo is a phenominally quick car and is generally robust, but it does have lots of problems with head bolts. So much so, that a website here in the US is loaded with data about these problems. One guy even went to the trouble of buying a set of bolts and taking them for destructive analysis. in all this reading I have always noted that they use torque figues all the way down. If you have time have a look at the Toyota website Supra / head bolts
Good thing about working on the small old cars in England when I was young. There's something about breaking a few studs that teaches a lot about the feel of metal. The early days of ally engines was a nightmare for a while. At least we could get a scrap engine for $15 !! Keep at it. :-)
This picture shows the hole that looks like it was plugged (at the factory) and then timeserted by me. You can clearly see a ring around the diameter which was quite a bit thicker before drilling for the sert. I was hesitant to sert this one originally because the plug looked to about the same size as the sert and was afraid that there wouldn't be enough meat left in that to securely hold a timesert. It seems the back section of this block (lateral centerline across cyl's 7 & 8 towards the back of the block) is very porous. The dimpling you see between the holes appears to be either trapped air or the aluminum cooled too quickly in this area when it was cast. When I tapped the hole (upper right) for the timesert you could see in the treads what appeared to be several airpockets in the the aluminum. The quality in the casting is markedly better before the back 2 cyl's.
11-26-04, 02:52 PM
It looks to me like there was movement somewhere and what looks like a bushing is really the original surface surrounded by an area that is worn lower due to metal fret. The outside diameter of the "plug" will correspond with the id. of the hole in the part that bolts up to it.
This is where the block 1/2's meet. As tight as the bolt was that came out of that hole I'd have to disagree with you and say there was no movement. The surface surrounding this hole is level with the block surface. If guessing, I would think that original hole was somehow damaged during manufacture and was repaired using a bushing. Good news is, I've reassembled the block last night and it went together flawlessly.
11-28-04, 08:25 PM
I don't know what reference you are using but there is a very clear and concise tensioning procedure for the Northstar main bolts called out in the factory service manuals. The pattern for tensioning the bolts is VERY critical to maintain the roundness of the main bearing bores. All the bolts need to be run down and then the inner bolts are all tensioned first in the pattern followed by the outer rows of bolts.
You should never, ever bring a fastener to the final tension/torque while all the others are loose. You are asking for trouble doing this and can possibly distort or warp the parts being clamped together and/or cause the fastener to strip since the load on it is so high as you try to bring the gap together on the first few fasteners.
11-28-04, 08:31 PM
There is a lot of info on why torque-angle is used for the bolt tightening specs in previous posts. Search using "torque angle" in the forum search feature and read my posts. The following was in a previous post so it seems a little out of context but should suffice...
Originally Posted by dkozloski
bbob, there are any number of high strength bolt applications in aircraft that are also tightened for stretch but usually it is something like rod bolts where both ends are accessible and can be measured with an outside mike or a go/nogo gauge. I read an article in Scientific American magazine written by an Australian scientist/engineer that said he had identified so many variables involved in tightening bolts by measuring torque that he thought the process almost worthless. My experience from measuring torque with a dial wrench while tightening bolts for stretch pretty much confirms his findings. Tightening bolts by "torque to yield", measuring stretch sonically, and other result based measuring methods is a great leap forward in my estimation. We just completed a repair on an oil-cooled German diesel engine that even the local dealer didn't want to mess with because they felt unsure of all the wrinkles involving fasteners and tightening. We just tackeled each bolt case by case and made up holding tools and measuring equipment as we went. So far everything seems fine.
Yea, you hit the nail on the head. Installing critical fasteners using torque alone is almost worse than useless...it is misleading in that you think you know what you are doing but you really don't because of all the variables.
That is why the torque angle method is preferred for modern critical fasteners. It is a means of replicating bolt stretch based on the pitch of the threads. In it's most fundamental form torque angle allows one to correlate a certain amount of bolt stretch to the thread pitch and allow a spec to be set, based on an angle to turn the bolt, that will stretch it the desired amount. The conclusion is then, that the stretch will be correct thus the clamp load will be correct regardless of the torque that it takes to achieve this.
The torque angle method is not infallible...i.e..if the threads start to strip as the "angle" is dialed in then the bolt will loose tension and the joint will fail. So, it too, has to be done carefully and if the torque required to assemble the fastener to the appropriate angle starts to fall off then the assembler has to realize this and correct the problem.
But, torque angle is still far far better than just torque alone. The "torque" part of the spec is designed to just take the "slack" out of the joint so that all the angle is going into stretching the bolt. The factors affecting torque go up astronomically once the bolt starts to stretch and the torque gets high so the initial torque setting is deliberately as low as possible to just take the slack out of the joint so that the torque will be as accurate as possible and then the angle achieves the critical load by stretching the bolt.
We measure bolt stretch routinely in blind holes be grinding the ends of the bolts parrallell and square and then with a sonic probe we can measure the length of the bolt by bouncing sound waves down the bolt and measuring the time between the echoes. This enables you to determine that the correlation between angle and bolt stretch to set up the spec.... The thread pitch...i.e... if there is a thread every 1.0 mm then turning the bolt will stretch it 0.50 mm..... helps to roughly determine the angle required and then actually measuring the stretch sonically confirms it. The production equipment is verified regularily with the sonic testing equipment to confirm the bolt stretch is there.
A bolt is simply a spring so the bolt diameter, material and metallurgy will determine it's spring rate so that we will know exactly what kind of load is available as the bolt stretches. The bolts coming into the manufacturer for all critical joints are certified vs. a load curve spec and routinely checked so that we know that the proper clamp load is being applied for the stretch that we can then monitor.
Actually measuring the clamp load in the joint is very very difficult if not impossible to do non-destructively so all the manufacturer can do is to control and monitor the factors affecting the clampload..i.e..the bolt spring rate and the bolt stretch.
11-29-04, 01:34 AM
One of the things I had been puzzled about for years...great post, as always thanks for your help.