[navdeep.singh]

Hi all

I have to finish the bores of a huge connecting rod which has bore dimensions of 250 and 310 mm with tolerances of about 20 microns and depth of about 140 mm.

Now the problem is everytime we finish the 2 bores with the required tolerances using boring bars, we get the correct tolerances after machining. However, after a day or two, the bore sizes go haywire. The bore becomes oval or has taper or both with tolerances varying over 50 microns. This completely renders our connecting rod to be rejected as the bearings do not fit in well.

We have tried the process of doing the roughing operation on one day and then finishing the bore in the next day so that stresses developed (if any) in roughing could be released. However we have not found better results.

Also it has been found that the bore sizes do not change beyond 5 microns if the bores are initially filled with weld material (welding is done on connecting rods whose sizes become unacceptable). So we assume that weld material being harder, does not cause the bores to change. However this cannot be done for all the connecting rods as it becomes very expensive and impractical.

The material used is Mild Steel with hardness of about 170-180 BHN and the bores are initially gas cut. We are using air instead of coolant for the boring operation as this gives somewhat better results than using coolant.

Your help in this regard would be highly appreciable since we have tried many alternatives but are not getting good results. Please suggest if we are going wrong somewhere.

[Geof]

.....The material used is Mild Steel with hardness of about 170-180 BHN and the bores are initially gas cut. ....

I think you need to normalize your parts before machining.

You do not say whether it is cold rolled steel or hot rolled steel but whichever it is once you gas cut you introduce stresses from the heating and cooling. When you machine the parts these stresses slowly relax and distort the material.

For normalizing you need to heat the part to around 800 degrees C and then let it cool in air.

[fizzissist]

Do you think adding a dose of shot peening would help with the stresses?

[RICHARD ZASTROW]

Had the same problem with large connecting rods for stationary gas V-12's. We bored them close, waited a few days then finish them to size by honing. BTW, don't use a flexible cylinder hone, the bore will bell mouth.

Dick Z

[Geof]

It has always been my understanding that shot peening was to improve fatigue resistance by inducing compressive stresses in the surface layer whick mask tensile stresses originating further down in the material. Anyway evenly shot peening something this size would be a real challenge, The bores are 12.2047" and 9.8425" so the rod must be several feet long. It also sounds like it is flame cut out of 5-1/2" plate so it must be full of stresses.

If the rough boring and finish boring were separated by several days, or even weeks, it may be possible to keep within tolerance. The problem is that you need to leave the absolute minimum for finishing otherwise the material will move further.

The reduced distortion after weld-up and rework is probably due to localized stress relief being caused by the heat from the welding. With extensive welding like this the material gets so hot throughout that it stresse relieves to some extent.

I think the only solution is proper stress relieving. If these parts go into service still containg significant stresses they will relax during operation and the bores will go out of tolerance and probably reduce the bearing life.

The convenient thing with normalizing is that it does not require controlled cooling. This means it is quite possible to build an enclosure out of fire bricks heated by several large torches. A full scale heat treating oven is not needed.

[navdeep.singh]

@ Geof: Your understanding of our design of the connecting rod is quite close. The gas cutting is done out of a 140mm thick plate and the centre distance between the bores is 3500mm, so it is a huge connecting rod.

Whatever you have talked about stresses is somewhat similar to our understanding of the problem but we were unable to understand the fact that these connecting rods did not distort when they were bored on a conventional machine where 2 rods were tack welded into a pair and then machined together. We have started using CNC machines only 3 months back before which machining was done entirely on conventional ones. Do you think we should tack weld them into pair and bore them together in CNC machines as well? Will it help in lesser distortion since the overall thickness will increase?

Also we try to keep minimum machining allowance during finishing. Normalising is something we could have to resort to if nothing works because it would take time and investment to implement.

Your suggestions have been really helpful. Do throw some light on the fact that they didn't distort on conventional machines, if you could. Thanks. :-)

@ Richard: Thanks for your response, however honing is an operation which cannot be performed on our machine so it would be difficult to incorporate that process. :-)

[Delw]

are you doing the exact same process on the cnc vs the manual using the exact same tooling? I am betting no this is why your having the issue.
people tend to hog the material out when roughing in a cnc( which you cant in a manual machine) this puts stress into the part.

also the way you clamp and how tight you clamp the vise could also be a problem.
to do it right get the material normailized, its best to hone them as well. other wise unless you have experiance your just chasing you tail

[RICHARD ZASTROW]

Delw, I think Geof will agree that conventional machines can hog every bit as much as a CNC machine. Some of those machines can take more. CNC has nothing to do with it. Machine construction and real power does.

navdeep.singh, Delw has a good point, is the process exactly the same? Try to repeat what you did to produce those parts including the clamping.

Personally, I would bite the bullet, do the normalizing.

How much time and money are being wasted reworking the parts? Remember, time spent on rework can not be recovered for other production which generates money.

Dick Z

[Geof]

If your setup and clamping is the same for the manual machine and the CNC I find it difficult to speculate that the CNC machining is the cause of the problem. Anyway if it was distortion arising from clamping you would be able to see it immediately the part was taken out of the machining.

I also doubt that the difference in machining could be the cause. And in my experience, like Richard says, manual machines can often hog harder than CNCs.

And I doubt whether tacking them together has any effect. Again, any distortion would be apparent immediately they were separated.

Are you sure the material is exactly the same. Are you sure they were torched out in exactly the same manner. I have had really annoying problems due to slight differences in material properties even though everything was still within specifications.


P.S. Delw; that would be some vise.

[christinandavid]

Do throw some light on the fact that they didn't distort on conventional machines, if you could. Thanks. :-)

Train your conventional machinists up and sack/demote your current CNC operators...

DP

[Delw]

Edit:

now I see its not metric(geoffs post) HOLEY CRAP thats a big bore. now that changed everything. including how your rough finish and clamp


Definately go with normalizing.

Forget everything I said bout the vise and the boring head, honing.


we have done many rings up to 50" in dia at the old shop I worked at in all types of material. one thing I learned is lots of unclamping and clamping to get the stress out of the tightened part is a MUST. as you put stress into the part where its being clamped at.

one thing I would do forsure and had to many times on big rings is.

Clamp
rough mill all of them (if Possible)
un clamp
then clamp again
semi finish mill all of them (if possible)
unclamp
reclamp
finish mill all of them( if possible)


the way you clamp the part will have a huge impact on how the part comes out.
no way would I clamp and rough and finish, you need to relieve it in between roughing and finishing operations.

I would also double check my fixturing t make sure you have it just right.

[navdeep.singh]

The machining process used is the same for both conventional and CNC machines. We are using the same angle plates which we used in conventional machines, the same rpm of the cutter while finishing and the same feed with the same coolant conditions.

The only difference being the amount of material being removed in roughing operation. While in conventional machines, the maximum cut being taken was around 1.5-2 mm per wall whereas in CNC machine we take heavy cuts of around 10-12 mm per wall. Also in conventional machines a pair was machined together whereas in CNC we machine them individually. These were the only differences as far as processes and clamping were concerned.

Although after the suggestions given by everyone here, I understand that we need to include the process of normalising after the roughing operations so that the connecting rod is stress free when it undergoes finishing.

As far as clamping goes, we have checked the bore sizes before and after unclamping but did not find any differences, which means that clamping is not the problem here. Lack of stress removal is the issue to be solved.

[Geof]

.....While in conventional machines, the maximum cut being taken was around 1.5-2 mm per wall whereas in CNC machine we take heavy cuts of around 10-12 mm per wall.....

Okay I guess I am going to contradict myself a little bit here. I am surprised you are doing such a heavy cut on a CNC, must be a big machine.

Anyway this is potentially the cause of the problem. When you whittle the metal away in small cuts it distorts a little bit with each cut as the stresses are being relieved gradually and each cut removes distortion that occurs following the preceding cut. This means that after the final cut there is not as much final distortion. Taking most of the material off in a single cut means there is much more distortion relieved in one go and this is takes it outside tolerance.

[navdeep.singh]

We have a heavy Double Column Milling machine with spindle power of about 22 KW so heavy cuts are not a problem.

However, as I mentioned, these heavy cuts are made only while roughing after which the connecting rod is allowed to rest for a day or two. Machining allowance of about 1 mm is left for finishing operation. As far as finish cuts are considered, we don't go beyond 0.5 mm of cut. So the final distortion happens with just 2 or 3 cuts of 0.4-0.5 mm. Do you think this will also induce the same amount of relieval of distortion to allow it to go out of tolerance??

[Geof]

..... Do you think this will also induce the same amount of relieval of distortion to allow it to go out of tolerance??

That I cannot answer directly. My experience has been working with flat material and I found it was best to approach the final size by many small passes rather than hogging 90% off with a single roughing cut.

Have you experimented by taking many small cuts with the CNC to replicate as close as possible the way you did them in the manual machine?


Also, but not directly related to the distortion issue, how are you doing the roughing in the CNC. With a boring tool or by interpolation? Obviously on the manual machine everything must have been boring. Have you ever tried plunge milling? It is possible you could reduce your roughing cycle time.

[navdeep.singh]

We haven't tried taking smaller passes in CNC machines yet but will surely try it now considering the implications that it could possibly have.

Also we do roughing by helical interpolation and finishing by boring bars. We have not tried plunge milling yet because we did not have the appropriate cutter earlier but now we are getting a deep shoulder milling cutter of 200 mm length which should do this job. I totally agree that the cycle time in the latter case would be much lower.

[navdeep.singh]

I have discussed with my authorities about the various aspects discussed here and as far as normalising is concerned, they are skeptical that the connecting rod might change it's properties after being treated to temperatures of about 800-900 degrees. Since it takes the maximum load in the machine, any change in the properties of the mild steel conn rod may cause it to break. Do you think the material properties would be affected in normalising?

[Delw]

The best way to normalize is to rough out part then normalize. if you just normalize solid bars/ slugs your defeating the purpose. any big bores should have a hole punched through them.

If you dont have anyone that can do heat treating/ normalize properly, your wasting your time and money. its very important that its heated correctly and cooled correctly, other wise you will have more problem than its worth.

Dont know what kinda steel your using Alloy wise, but with a quality alloy you wont have any problems. if your using junk steel with lots of impurities than again your wasting your time, and this also could be a problem as well with your machining problem your having.


Delw

[navdeep.singh]

We are using hot rolled mild steel plates for our connecting rods with hardness of around 170-180 BHN. It is not junk steel that we are using.

[Delw]

its worth a try. I send out 9310 for normilize and some other alloys, never tried hotroll.

I can tell you the alloys I send out the normalize process is the only way to go.
its called normalize harden and temper. usually to around 31-33 RC, we usually rough out first and if there is any holes in it they always get roughed out.
not to mention it makes cutting some much nicer.