Fully Threaded Rod Versus Threading Ends Only

Thread the ends only. The cross-sectional moment of inertia and the radius of gyration will take too big of a hit when root diameter is compared with full diameter.

Yes, I know the root diameter is smaller than the full diameter of the threads. That was stated in the problem. But the comparison is with threaded ends AND the remainder of the rod approximately the same as the root diameter, NOT the full diameter, versus continously threaded rod.

A simple analogy: airplane flying wires (seen on biplanes, most typically) have long slender aspect ratios, and cannot handle any compression loads. The never see the sort of accelerations described here, and will see far fewer cycles.

But they ARE life-critical, so there is great incentive to get it right. Only the ends are threaded, and it's done by rolling.

Spokes on bicycle & motorcycle wheels carry tensile-only loading, and the hub end is turned at 90 degrees, eliminating any chance of taking a bending load. They, too, are only threaded at the end. If all-thread were cheaper, would that happen?

Roll threads on the ends, only.

"Roll Threads on end only" does not explain anything. I asked which is better and WHY?

<UNSUBSCRIBE>

You get what you pay for.

Sorry, I did not ask for an alternative solution. I asked which of the two approaches described would yield better results. And your point about machining the rod is exactly the point I tried to make with my boss.

Let's make it really expensive and flute the un-threaded portion. Just a thought.

Agree with Lyn. Rolled threads allow using stock smaller than cut threads require. Consider using a grade of steel that has the best properties for fatigue.

I already stated we would use rolled threads and this issue of which grade of steel to use is not germaine to the comparison of the two design approaches.

You did not read the entire post carefully. This is not an automotive piston connecting rod. Those are subject to non-axial loading due to high rotational/non-axial inertia as one end is connected to the offset pin on the rotating crankshaft, which dictates their beefier construction. I did not ask will it work or not work, just which approach would have better fatigue strength, given only axial loading.

I'm with tcmtech on this one. accelerating the rod alone induces a stress of 12ksi at the middle of a 24" piece and twice that at the end where the force is applied. I don't know how much mass it has to push in addition to its own weight. The kl/r ratio would be a problem with a small diameter rod.

Better from a slenderness ratio perspective, would be a larger diameter rod, perhaps turned down if necessary and rolled at the ends

Regardless of "how much mass it has to push besides its own weight" (which is moot since both alternatives must do the same job), given axial loading only, why would the stress at the ends be twice that in the middle?

Gut thoughts:

1. A 27 in. long x 1/4 - 1/2" diameter rod with 'high axial' compressive loads is a column and appropriate design will have to be considered. The 1-5/8" ID tube adds no lateral support until buckling is well along.

2. Leaving threads on the center portion of the rod, besides the stress concentration factor, adds considerable weight (which must be accelerated at 'several thousand G's', causing additional axial load) while adding negligible strength.

3. Unless this machine is a cross-head design, the connecting rod will see considerable bending loads from the lateral motion at the crank end vs momentum of the rod.

4. Agreeing with tcmtech, a tubular rod will give a higher strength (bending and columnar) to weight ratio than a solid rod.

1. Correct. I was thinking of adding lightweight support discs at intervals along the rod.

2. UNF thread form provides much shallower threads than UNC. I do not believe the added strength is negligible, but rather a trade-off for the small added weight.

3. Again, not a piston connecting rod, so no crank, no lateral motion. Axial motion only.

4. Again, not asking for alternative (tubular) solution, just a comparison of the two approaches presented.

The rod is much too slender, even if you can be sure the motion is purely straight line reciprocation. The largest diameter tube you can fit in the space would be better.

Try saying this a few times in rapid succession:

One smart fella, he felt smart.

Thank you to all who posted helpful comments. Actually, the slender rod inside a hollow tube is my boss's idea. The original design was to be a 2" OD Tube made of Al/Ceramic composite material, but cost was quite high and my boss decided it could be made of PEEK/graphite fiber material, but that was not quite strong enough. The 2" OD shaft is guided by rollers (in the testing area) and by a hydrostatic linear bearing where it comes out of the crankshaft box.

I agree that the design as-is is probably not workable, but was looking for some resolution only of the two choices my boss and I differed on. I agree we need to beef up the rod, but added weight would be very problematic. I am now looking at an alloy steel tube, .625" OD and .450" ID with 1/2-13 UNC studs welded into the ends. This gives only a small increase in weight over the solid rod design. Even that may not be enough, but it is a start.

"Failure is not usually an option... But it is often a design criteria!"

In your thread options, how about rolled threads vs machined cut threads, and then there are thread profiles to consider and the nut types, with or without nylon inserts, split pins etc etc.

I know this is most likely not even worth posting but... Other that the threaded ends it sounds like valve push rods used in an engine. Just a different direction in thinking was all.

First, one has to design the rod for the worst stress condition, which appears to buckling (the containment tube is irrelevant here, it is the first, microscopic crippling effect which initiates an unstable failure). This will dictate the diameter subject to a design check for fatigue, which may be material dependant.

Because of the need to have an adequate moment of inertia, the diameter is likely to be greater that is needed at the threaded ends, so fully threaded rod would have to be larger. It is likely that the ends could be reduced but rolled threads exceed the effective diameter of the rod.

There's nothing wrong with pushing on a screwed rod but its stability becomes the design criterion.

Threading is a mental fixation and a stress concentrator, not a valid design criteria. Connecting rods evolved into the present form for good reason(s). That is a starting point.

If it is true, that there is really no sideways movement, intended or not (and that is a big IF), a thick metal sleeve with a forced lubrication may contain the incipient buckling of the rod or tube. Metal to metal rubbing cannot be allowed.

Congratulation. You replaced a traditional connecting rod with a fiendishly difficult lubrication / cooling scheme. Good for test equipment, that's all.

Forget welding. Low grade, much variability.

Since you have already stated that various materials have been considered for this task, have not provided details of the loading on the rod, and stated that buckling and fatigue are concerns, I suggest you investigate other materials for the rod such as titanium (perhaps 6Al4V). Its lower density (compared to steel) will substantially reduce inertial stress thus substantially increasing its fatigue life. After all, titanium is what we use for long life connecting rods.

Both designs will work, It all depends on how much money you want to spend and how often you want to replace the rod.

"Conventional wisdom". Define both of those terms and you will see the answer clearly.

We need to know the compressive load to calculate the Moment of Area needed to avoid crippling,