Questions regarding thread diameter and pitch for special design case with limited thread length

Going to a smaller thread will only make it weaker. The threads will pull out easier under the resulting tension component and the smaller diameter where it meets the 3/8 diameter is where it will likely fail. Extra fine threads will not help the pull resistance in the tension component but the the larger minor diameter will help somewhat at the junction with the 3/8 diameter. I would not go crazy with and extra fine thread unless you happen to have the tap and die. Since you are limited by the rod material the best thing you could do is to drill and tap the rod as deep as possible and use a extra high strength set screw. The wall will be only 1/32 but its material is the right place being farthest away from the neutral axis. You might want drop down to a

1/4-28 screw to give you a little extra wall.

Hi Tomcas,

Thanks for your feedack,

A 1/4-28 thread would probably give me one more thread than a 5/16-24 thread. Also, as you point out, the 1/4-28 would also give a little extra wall or contact area where the necked down portion of the rod makes contact with the support.

The main reason I considered 32 threads per inch is just to get more threads on the short length.

I hate to take the extra step of putting the set screw in (or possibly a dowel pin) just becasue it's an extra manufacturing step.

Perhaps the 1/4-28 thread would be better than the 5/16-24, just because of the extra thread ?

Thanks again for your feedback. John

It sounds like you just want to believe somehow a thread choice will make up for lack of material and/or tensile strength of that material. It doesn't work that way.

Re best pitch, I agree with tomcas. Put a nut on it. Read on...

Even if the "necked down" part is machined, you don't have much bearing area in compression on the bottom of the rod and compression force there is high for given moment because the radius is small. So only a bit of yield there transfers all load to the thread.

Stress on the thread then varies with distance from the axial horizontal plane with some parts stressed more than others.

Thread the rod a bit more and put a nut on it, then torque the nut down tight against the mount. That gives you a larger bearing area in compression resulting in more evenly distributed stress on the engaged threads, thus reducing maximum stress on the thread.

If you look at the rod as a lever, the fulcrum is the bottom edge of the

3/8 section that bottoms onto the vertical steel support. When the vertical load is applied to the end of the rod, say 5.5" from the fulcrum, it is resisted by a horizontal axial load only 3/16" (.1875") from the fulcrum (its not really 3/16", but you get the idea). Doing the simple maths, the horizontal force is then 5.5/0.1875 = 29 times the vertical load. If you weld a 1" diam steel collar to the end of the rod where it buts against the steel support, the horizontal load is 0.5" from the fulcrum and is therefore reduced to 5.5/0.5 = 11 times the vertical load. This way the horizontal load on the thread is reduced considerably.

You need to determine which one of those failure modes is happening.

If you cannot change the design to put a larger diameter where the part mates to the vertical surface, then get rid of the turned down section. Thread the part 3/8-24.

Jim

You have a structural design problem. Knowing the load that will be applied, you need to design a joint that will withstand the stress. This calls for a little engineering, either by yourself or a professional. Bugs

John, It really isn't an extra step, since you do not have to turn down and thread the end of your rod anymore. Instead, you drill and tap for the set screw, a very easy task. Trying to thread up to the shoulder, especially with such a short stub to thread, is an exercise in frustration and won't produce threads to the shoulder. And a grade 8 set screw will give you the highest thread strength you can easily come up will.

Dennis

Hi everyone,

Thanks for your replies.

I cannot thread the rod through the support and/or put a nut & washer on the other side of the support, or change the dimensions to larger parts. The "support" is actually a round bearing housing with a flat milled on it where the rod threads in. The threaded rod is basically a lever that oscillates the housing back and forth. Everything is in a very tight space, and the rod threads in right over a bearing, and practically bottoms out on the bearing (this is why I cannot go deeper and use a longer thread).

I figured I would just tolerance the thread length so it does not bottom out on the bearing OD. Moving the needle roller bearing out from underneath the rod does not help much since everything is cantilever mounted, and there is still a shaft underneath the rod even if the bearing is moved (you only get about a .05" addition to thread length). Also, this would put the bearings closer together, and I want to keep them spaced further apart if at all possible.

I also thought of forgetting about the thread length tolerance, and threading the rod into the housing before the bearing bore is drilled. This way, if a little thread extends into the bearing bore, it will be machined away when the bearing bore is drilled. The problem I see here is that if I have to take the rod out and put it back in, it will probably not go back in right.

Welding the joint seems undesirable (it will deform the bearing bore) . Heat would also deform the bore in the rod itself. I also don't have the space to add any type of collar as garry suggested.

Based on all of the feedback I have gotten from this and other forums, it seems the size of the lip where the rod necks down is fairly unimportant, and I should keep the OD of the threads as large as possible. It also seems a course pitch thread is better than a fine or extra fine thread, even though there may be less threads over the .183" length.

I misunderstood what tomcas suggested with the set screw, I can see it would not be an extra step, and may be a better choice than machining the shoulder since I would have threads along the full .183" length. I would probably use loc-tite to keep everything from backing out or loosening. I am not 100% sure it would really be stronger "overall" though, becasue then the weak point looks like it would be the thin wall left on the rod after it is bored through and tapped for the set screw.

The way things are now, the existing 6mm OD hole in the center of the rod does not go all the way through, and the drill point stops about .12" from the shoulder where the threaded portion of the rod necks down. Drilling the hole all the way through and using a set screw looks like it could weaken things overall, (i.e., the threads are stronger, but the weak point then becomes the thin wall of the rod where it connects to the housing ) I don't know that for sure though. If a 1/4" OD set screw is used, I would have a 1/6" wall remaining on the rod where it connects to the housing, but it does not seem like much.

How close could I realistically get the threads to the shoulder where the rod necks down? I guessed that with a 5/16-24 thread, I would loose one thread at the shoulder and wind up having three useful threads.

If I have the space, I guess the best solution may be to do away with the necked down portion, and try to thread the OD of some 3/8" drill rod. The rod threads into a 1" OD round, but it is right at the edge, i.e., the center of the 3/8" OD rod is only .236" from the edge or end of the housing.

When I said before that the load tends to pry the rod out of the support, I did not mean that the load actually caused the joint to fail, but just that that is how the load acts on the system.

As I understand it, you can have cumulative stress on threads in a cyclic application, so you can test it, it can work fine, but fail down the road. I have no way to simulate a few years of use in a short time, and that's why I made the post on the board to get feedback. The loads should be fairly moderate, say 10 to 15 pounds but the thread is so short I wanted to get some feedback so I can try to do the best thing right from the start. Even though the loads are moderate, the short thread length just worries me.

Thanks again for your help. John

Correct.

It also seems a course pitch thread is better than a fine or

Wrong. Finer is better. It has a larger root diameter.

Now that's a good idea. Extending the thread way past what is needed will move the stress concentration of the last tread out to a lower stressed area and loctite will hold it in.

there will still be a stress concentration at the point where the wall of the vertical support coincides with the one thread root of the rod that occurs at that surface. If it fails in shear, that is the location where it will fail.

If you can, keep the threads completely inside the vertical support and relieve them to the root diameter where it passes out of that material. From what you say though it sounds as if you don't have enough hole depth to do that.

In that case you will have to rely on the larger root diameter strength alone. Remember cross section goes like the square of the diameter. This might be enough.

If it fails by snapping at the point where the threads run out into the exterior portion of the rod, then you can improve matters somewhat by relieving that section down to the root diameter for a length of about 5 or 10 threads - but you MUST put gentle radii on the reliefs. I like do do this with a ball-nose end mill in the toolpost.

Jim

Why not just cut a J thread.

He could do that. The main stress concentration will still be at the thread root that lines up with the vertical surface though.

Most folks will have trouble getting the specific thread form like that, but it would probably increase the strength by about a factor of two or so,

*if* it were a rolled thread. Lathe cut threads are inherently not as strong so one has to keep the overall assembly in perspective here:

Most times very high strength fasteners are used in engines where rotating mass has to be kept low, or in aircraft where overall weight is a major concern.

Basically any time a simple machine is designed such that the extra strength of a special fastener is deemed neccessary, one needs to re-evalute the design and in most cases simply using a larger fastener or re-designing the connection is the best way to go.

At one time bmw made their flywheels bolt onto the back of their crankshafts using special M10 bolts. Waisted, high strenght steel, relieved shanks and all. Special torque values and all.

Those machines were *still* prone to spitting the flywheels off the damn cranks. Their decision for the next subsequent years models was simple: they went to an M12 bolt that was pretty plain jane. And the flywheels stayed on from that point.

Mostly the folks who *need* high strength fasteners know how to do the engineering so they can be used properly. Those who don't are, honestly, better off 'using the next size up.'

Jim