Internal Threads
 

Recently I've been thinking of the possible applications of internal threading in building the robot. I think it may help lower the robot's overall weight, losing the weight of nuts, and it would be much easier to assemble, being able to use only one hand and having no nuts to drop.
The thing is, I have no idea how to make them without spending a week preparing them.

What do you think the advantages or disadvantages of internal threads?

Also, did any of you ever use internal threading and has some insights about making and using them?

Re: Internal Threads
 

Bar,

Used properly, they are great.

Internal threads can be misleading. What are you applications are thinking about converting? What fabrication methods/material types do you currently use?

Yes, these can appear to save weight. However, inorder to get the same strength and results as a nut would provide, you often have to increase the metal thickness. These also don't provide sufficient locking features, but you can insert locking heli-coils to solve this (as well as a few other methods).

I don't want to discourage from this, but know that it is hard to tap (method of creating internal threads) a 1/4"-20 hole into a piece of 1/8" material.

There are some other "lightweight" alternatives out there.

Re: Internal Threads
 

As a general rule of thumb, you need three threads in the material, with at least one exposed thread beyond that. That means that for 1/8" thick material, you need a pitch of at least 24 to properly fasten. That's a lower limit on TPI, and you'd probably want to be higher than that.

I've never used a nut in a piece 1/4" thick or more on a robot. At that thickness, everything gets tapped. At less than that, it depends on the required strength. If the joint is in compression, tapping the material is usually sufficiently strong. Same goes for low load holes, like control system mounts, etc.

If you need extra thread strength in thin material, look into Pem Fasteners. They are pre-threaded insert fasteners, sometimes called "captive nuts" on web sites like McMaster. There's a whole variety of types for different applications. Most of them are designed to be clinched (read pressed) into sheet metal.

As for making the threads, if you plan during design and make the appropriate holes a tap hole for the appropriate screw, then it's actually pretty simple to fabricate. Get handy with a Gun Tap (aka spiral tap) and a properly clutched drill and you can churn out 20 holes a minute. Just make sure you properly lubricate your tap, especially if you are tapping aluminum. Personally, I'd recommend hand tapping any steel.

Speaking for my team, our entire chassis design doesn't have a single nut on it. It's all Pem fasteners (Pem nuts to be exact) and tapped sheet metal. If you watch your tolerances, pay attention to the difference between tap and clearance, use loctite (a drop of 242 now prevents a lot of loosening later) and account for some special considerations on the Pem fasteners (you'll find that in the datasheet)...it's a pretty smooth design with fewer maintenance operations than a nut-bolt paradigm.

Re: Internal Threads
 

I really love tapping (see my signature)! That said, I have heard some different rules about it. First, I've heard 6 threads, which would make 24 tpi the largest recommendable in quarter inch material. Also, tapping on a CNC mill can work, but can result in a tap getting broken... I would certainly recommend hand tapping for most things rather than going a marginally faster but more risky route.

In regards to PEM nuts, I would be more cautionary. First, some require a special tool to install, so most teams would need to make an investment to get into the system. Additionally, they can get "spun" if installed incorrectly or too much torque is applied. Like all fasteners, they are great for some things, but have their drawbacks. I still prefer rivets for frames and primarily tapping for manipulators.

Re: Internal Threads
 

on the subject of attaching things with screws without using bolts or t-nuts anyone else ever take a look at or use anything like "nutstrip" or "attachment blocks"?

Re: Internal Threads
 

Having used these guys for 1/8 sheet back on 1647, I can tell you they held up extremely well. We used approximately 10 of them to mount to our bellypan for our arm and under the loads the arm was forcing down through to the bellypan, there was no movement from them whatsoever.

Re: Internal Threads
 

For a demo project in 2005 and later to build the competition robot for 2006, 931 used PEM-type 1/4-20 press-in nuts. Several boxes of them. I got them from McMaster. You can see several instances here.

Tip: use a good sharp drill, or better yet, a good punch to make the holes in the sheet stock. Leave the hole edges sharp (i.e., don't deburr) because the press-in nuts clinch to the sheet by distressing those edges.

Re: Internal Threads
 

Another option for internal threads is using threaded standoffs instead of spacers for gearboxes and other parallel plate applications. We started doing this on 2168 in 2011 and it has been a positive experience. No more looking for spacers during repairs.

Here is an example part from McMaster.

Also for our bumper mounts we used some Riv Nuts (also from McMaster) installed in the frame, and some large thumb screws as fasteners.

Re: Internal Threads
 

First of all, thanks to everyone for the quick replies :)

Secondly, most of you mentioned Pem nuts. From your posts and the pictures, they seem pretty permanent and hard to remove, and we usually find ourselves making significant changes around day 20...
How easy (or hard) is it to make changes when using this?
How much planning ahead do you need to use them correctly?

Re: Internal Threads
 

The number I know is 3 complete turns if the threaded material is as strong as the fastener. Because high-grade aluminum can be about 1/2 as strong as an alloy steel bolt (this is subject to wide variances) 6 turns would suffice.

Now, if the application actually requires the full strength of a selected fastener is a whole different matter...

OP - you might also want to look into rivet nuts, they can be assembled easily and quickly with basic hand tools. I don't want to start a "rivet nut vs pem nut" debate, suffice it to say that they each of their uses, so consider your choice carefully.

Edit: There are also anchor nuts and weld nuts that might work well, depending on your application. Some anchor nuts can be purchased with nylon locking elements and/or floating threads, making them very useful where small adjustments might be needed.

Re: Internal Threads
 

And thus we risk diving into an ongoing debate.

The guidance I have received from a number of sources is that 3 threads is the minimum to get the appropriate geometry for the threads in the hole. If you think about what having a thread or half a thread means...well that's not very reliable.

As for strength, there are definitely places on a robot where the strength of steel is needed, so there steel should be used. But for joints in compression (like say the corners of the frame, or a gusset plate in the corner of a frame), you probably don't need the full strength of steel in the fastener. Most of the time, the load will be carried by the interface between the two pieces being joined, not by the fastener itself (assuming it's properly tightened). Joints in tension are another story, where the load IS carried by the fastener.

And, as I mention, the required strength of the connection is a factor. The design guideline we use on my team is that all of the sheet metal structural connections get PEM nuts, so that we have a steel nut with a steel bolt for strength. Plates 1/4" thick and over are tapped without the PEM nut. We do, however, tap 1/8" aluminum sheet metal for control system mounting holes (jaguars, PDB, sidecars, etc) using #8-32 or #10-32 cap screws.

Regarding fabrication, if you get the correct PEM nut (that would be a self clinching nut in the S, SS, CLS, or CLSS series...each series denotes a different heat treatment or finish), they can be installed using a press or even an impact driver (though I wouldn't recommend the latter). We've used 1/4"-20 fasteners in the past, but the nuts are substantially harder to clinch than the #10 or #8.

There are other very interesting options from PEM. For example, they have PEM studs, which are threaded connections that stick out of the sheet. I find these quite appealing for the aforementioned control system mounts. There are also floating fasteners, which are intended to account for variations in manufacturing. They basically "take up" the tolerance, and allow for a slightly fudged hole to still properly fasten. For example, using a floating fastener is probably a good idea if you are going to bolt down all four corners of the PDB.

Yes, there was a time when I thought lock nuts and cap screws were all there were to fasten things. Then, I did what good engineers do...sought more experienced help. That's lead to a much more enjoyable build and less maintenance at competition...which is good for all.

Re: Internal Threads
 

Thus why nuts generally have 3-4 threads. Funny how that works :)

/thread jacking

Re: Internal Threads
 

Right. Key takeaway: thread strength is not additive. The first couple of threads take more than half of the bolt load, with each successive thread contributing a lesser and lesser amount.

Re: Internal Threads
 

I love debates! You get more information than you signed up for.

Again thank you all for the comments.

Do you ever tap wood or plastic? Or is that the wrong way to go?
Also, what tools do you use to tap? I don't know any, I'm afraid to say...

Re: Internal Threads
 

Specifically addressing this...

Making changes isn't all that easy, particularly if you're off by less than the diameter of the PEM fastener. When the error is that small, you can't make an appropriate hole to get the fastener to clinch.

So, yes, a PEM nut is pretty permanent. Using them requires planning. But, spending a respectable amount of time in Inventor (or SolidWorks, etc) getting all the constraints and tolerances correct is really all you need. And, if you have to make a change on the fly, it's not much more work than drilling a hole for a nut-bolt connection. You just have to make sure you get it right the first time.

Keep in mind that it's not an all-or-nothing decision. You'll use the same bolts in the PEM nut as you would in a regular hex nut. And, you can always use the hex nuts in some places and PEM nuts in others (just make sure you have the corrrect size hole in the correct location). I'd recommend that you start with PEM applications in good, static, well defined things, like chassis construction before moving on to the potentially more complicated parts.