pic: 3647 Frame

Cannot wait to finish this…

Super excited to see your finished robot, Mark! Hope you finish up that CAD soon :slight_smile:

Love the holes in the square extrusion, we did this by hand last year on our robot and it was completely worth the time.

Did you have yours machined or was it manual?

I cannot wait to see the finish bot. heh…

What CAD :wink:

This frame was machined at a local sponsor’s shop by me. The grunt of machining holes in rails was done on a 2 Axis CNC (Prototrak) converted Bridgeport, with the CNC positioning I was able to bust the rails from 5-10 minutes each, the machine was big enough to do any rail in one path and also had an power draw-bar for super fast tool changes.

The debate on whatever doing a hole pattern in every rail is worth it or not is one that my team is still kinda spilt on. A lot of people have told me that I could have “saved a lot of time” if i just drilled the holes I “needed”. I however believe that having a standardized bolt pattern to be a life saver and make general construction a breeze. And allows for fast changes to the design, plus doing them on the mill makes sure they are all square and dead on, and with our machine abilities it wasn’t that much time.

I’m a believer in the standardized bolt pattern and we did them with hand drawn lines and a cheap drill press. One of my students did 1700 holes over a 3 day weekend. I tried to get her help but she said it would be faster if she drilled everyone and just had other team members mark them for her. We hand cut the gussets on a band saw and drilled them with printed templates.

Some of the bars can be seen in this post.

We ended up needing the holes very often for design changes. We added a separate section to our ball elevator in about an hour because we already had all the holes ready. I’d really like to see more teams use this construction method.

Are you using all 1/8" or is most of that 1/16"? We did all 1/16" and only had a few poles bend due to ramming into the bridges but otherwise it worked out perfectly.

Wow… I can’t even think about doing all those holes by hand… #SpoiledByCNC

The 2x1 is all 1/8th and the rest of the structure is a mix depending on what joints require more strength, we did have some bending issues with 1/16th last year and the bridge as well.

Makes one wonder why someone doesn’t produce and sell 1x1 holed square tubing. Closest product I could find was 1.5" x 1.5" “Holey Tubing” sold by 80-20 guys. Are you listening Andy?

The point of this is process is that its very cheap. We are able to get 1x1 square tube with 1/16" wall for about $14 for 21 feet. If AM or VEX could sell something near that price I would be all over it, but I can’t imagine they can do that.

Dang…if that girl’s up for some more speed holes, your team could make some serious cash selling that stuff…

Looks beautiful!

I still don’t understand how people are thus far in their robot build. We are still desiging in CAD, but have sent out parts to get water-jetted. Maybe We’re just behind… but oh well!

Thanks for the compliment, she won’t be repeating the feat this year though. Our design doesn’t call for much drilled extrusion and we have a sponsor with a laser cutter.

But for teams looking for a nice way to build robots in a small shop, I would highly recommend this method. The holes don’t have to be perfect to be useful, just close enough. We used a 13/64 drill bit so it’s a clearance hole for a #10-24 and small enough for a 3/16" rivet. We did it all on a cheap Ryobi Drill Press. Most people don’t believe it when we tell them it was done by hand.

Like most hand produced but repetitive tasks this job can be improved with a jig for quick alignment of the holes. Here is one example of how it could be done, no markings necessary.

We did think to make a jig, but all our attempts ended up far more inaccurate than hand marking and center punching.

The problem with a jig for this is tolerance stackup. If you’re drilling 50 holes 0.5" apart you’ll be an entire hole off by the end if your jig if consistently 0.010" off. If you’re marking every hole by hand, you can be off by 0.010" on each hole and it won’t matter much because that’s 0.010" off the true dimension, not the distance to the last hole.

If you want a really long setup, you could make a similar jig and use a previously drilled bar as an indexing plate. As long as you use the same bar as your indexing plate, all the copies of it would be pretty similar.

My team ran in to this exact problem last year and arrived at the exact same solution. We had a small CNC mill that we used to make a template and then had students use it on a drill press. We didn’t just use the mill for the entire operation because it is much easier to train a freshman to use a drill press than it is to train them to use the mill, which only 3 people knew how to use to begin with.

How hard is it really to train a willing and motivated student to use a basic mill, to do basic operations?

It took me about 15 mins to get people to start doing bolt patterns on our in shop mini mill which has an DRO. And because all they had to do was turn a dial to .5 increments they were dead on precise. I used the CNC Bridgeport for the longer rails (anything over 11 inches) due to the extra travel and speed.

Quoted for truth. If you can’t do operations like this on the mill, why have a mill?

Fairly difficult when the mentors and experienced students who would otherwise be doing the teaching were still learning themselves.
The mill was a fairly recent acquisition that the team had not yet had time to learn how to use efficiently, so we made do with the experience we did have.

I agree that using the mill to drill the holes would have been preferable, both in the interest of accuracy and in the interest of teaching skills. As always, it was a matter of utilizing the resources and experience available.

If VEX can make beautiful 7075 gears for a few bucks apiece, I’d bet they could do some box tubing with holes in it for a great price too. They essentially already do this, with slightly different products.

I would normally agree, but have you looked at the price per foot for VEX metal? They obviously haven’t moved manufacturing overseas for these items but I am not sure they can get down to the $1 per foot range even if they do that.

It’s currently $40 for 6 feet of 1x1x1 C channel](http://www.vexrobotics.com/vexpro/chassis-and-framing/vexpro-frame.html). (This is thicker sheet metal and not extrusion but I would think it’s comparable.)