FRC pocketing

[GeeTwo]

Originally Posted by MrForbes

There are different ways to look at pocketing... another way is to see it as a mostly unnecessary process, required only if you designed something with the wrong material, or wrong shape.

But I'm lazy and cheap, so take this with a grain of salt

Originally Posted by JVN

Jim,
Please provide some guidance to inexperienced teams on how to always design with the right material and right shape.
Otherwise, perhaps withhold advice like your previous post as it might send someone down a path they aren't prepared for.

"Mr Forbes says we don't need to do pocketing!" *140 lb robot*

148 loves pocketing / trussing. We do it on everything. Maybe we're just bad at designing with the right material & shape.

-John

Originally Posted by TheOtherGuy

It's probably because 148 isn't cheap or lazy

One simple alternative: use thinner material without pocketing. I've seen a lot of pocketed designs in FRC that could have been made simpler and cheaper with practically no effect on strength (although, oftentimes can be made stronger) by using thinner material.

The difference is easily explained by analogy to bones. While mammal limb bones are simple tubes with filled with marrow, they do not have any truss work within them. This is because mammals are not pushing the envelope on the tensile strength of calcium and phosphate salts as hard. Dinosaurs and birds however, push that envelope. Land dinosaurs came in sizes much larger than mastodons, and birds need to shave every available gram in order to improve performance in the air, so their bones feature (irregular organic) truss work. Here's a web page comparing the structure of human arm and bird wing bones.

[EricH]

Or use a completely different material. Wood is nice for certain applications; PVC has its uses. You can't forget about fiberglass/carbon fiber (not the same material, but I'm lumping them together) despite the extra precautions needed for them.

In short, here are some ways to avoid pocketing/swiss-cheesing:
--Change material. Aluminum fasteners and gears instead of steel ones, for example. PVC structure can be heavy, but try building it out of aluminum sometime.
--Use a different design.
--Use thinner material that doesn't need to be pocketed.

There are also some other tricks not having to do with pocketing or making lots of holes.

[MrForbes]

I'll provide a little history behind my comment, and why I think JVN's hypothetical "Mr Forbes says we don't need to do pocketing!" *140 lb robot* is quite humorous.

I started working with team 1726 on the last weekend of build their rookie season. My son was on the team, and he asked me to come in and help them get the 140 lb robot down to 120 lbs, and they only had a few days to do this. So, I helped do a lot of pocketing and we got it to 119.9 lbs and all was well. Of course, I got hooked on the whole FRC thing... The next ten robots the team built, I was around at the beginning of the design process, and we didn't ever build another robot where we had to chase weight by cutting material away like that. Instead, we kept track of weight from the beginning of the design. And we didn't use a material because "that's how robots are made"...we kept open minds, and used materials that met our requirements of cost, local availability, weight, strength, stiffness, ease of fabrication, suitability for that game, etc. It turns out you can use a wide variety of materials to build robots. We've used fiberglass, wood, steel, aluminum, polycarbonate, etc over the years.

Mainly, I see the extensive use of CNC material removal as kind of wasteful. This mostly has to do with the fact that I'm cheap and lazy.

If you have the resources to purchase extra material and then cut it away, then go for it!

[AdamHeard]

Originally Posted by MrForbes

I'll provide a little history behind my comment, and why I think JVN's hypothetical "Mr Forbes says we don't need to do pocketing!" *140 lb robot* is quite humorous.

I started working with team 1726 on the last weekend of build their rookie season. My son was on the team, and he asked me to come in and help them get the 140 lb robot down to 120 lbs, and they only had a few days to do this. So, I helped do a lot of pocketing and we got it to 119.9 lbs and all was well. Of course, I got hooked on the whole FRC thing... The next ten robots the team built, I was around at the beginning of the design process, and we didn't ever build another robot where we had to chase weight by cutting material away like that. Instead, we kept track of weight from the beginning of the design. And we didn't use a material because "that's how robots are made"...we kept open minds, and used materials that met our requirements of cost, local availability, weight, strength, stiffness, ease of fabrication, suitability for that game, etc. It turns out you can use a wide variety of materials to build robots. We've used fiberglass, wood, steel, aluminum, polycarbonate, etc over the years.

Mainly, I see the extensive use of CNC material removal as kind of wasteful. This mostly has to do with the fact that I'm cheap and lazy.

If you have the resources to purchase extra material and then cut it away, then go for it!

I think the points you make here are valid for many teams, but there are two other points relevant teams should be aware of.

Each team has unique resource, so for many teams it might be more efficient at the team level to use 1/4" plate and pocket it every time than try to figure out lighter methods (ideally some compromise can be found here).

The other point is that there are some games where the top 10% of teams or so that try to do it all (or some large subset of doing it all) and to pull this off most of those teams (based on the above point) can't hit weight without pocketing.

To sneak in a third point, there are reasons that coming way under weight can be valuable (namely faster acceleration and less battery use).

[Aren Siekmeier]

These posts have pictures of some good examples of pocketing.

http://www.chiefdelphi.com/forums/sh...stcount1387268
http://www.chiefdelphi.com/forums/sh...stcount1396742
http://www.chiefdelphi.com/forums/sh....php?p=1308653

It's often used because a flat surface is needed (or is convenient for fabrication) or the thickness is needed for bearings or geometry, and then you cut away material to get a truss structure between the important (load bearing) points in the part. Reduces weight, can ease maintenance, and also looks kinda cool.

[Mark Sheridan]

Originally Posted by AdamHeard

To sneak in a third point, there are reasons that coming way under weight can be valuable (namely faster acceleration and less battery use).

or forgetting a key feature like a second carriage. D'oh! Being underweight was really helpful for 3476's design iteration.

[Al Skierkiewicz]

Let us not forget the practice of pocketing the back of bumpers to fit over protrusions in the robot frame. This practice was essentially eliminated with the FRAME PERIMETER language and "minor protrusions' limited to 1/4" entries in past rules. 2016 rules may differ, please be sure to read everything in the robot section, twice!

[GeeTwo]

Originally Posted by EricH

Or use a completely different material. Wood is nice for certain applications; PVC has its uses. You can't forget about fiberglass/carbon fiber (not the same material, but I'm lumping them together) despite the extra precautions needed for them.

And yet, each of the materials you list (and probably every reasonable FRC construction material) can produce a better load-to-weight ratio when under compression through the use of pocketing than with solid convex parts.

Tubing, angle, c-channel, and many other extrusions and sheet metal folds are simply an example of pocketing in one dimension. Monocoque is another great example of sheet metal one-dimension pocketing.

Did you ever really look at a cantilever bridge? Here's a neat example. Can you even count how many levels of pocketing appear here? (This is a 1930s railroad bridge around which another bridge was added much more recently; I grew up five miles away from it and still cross it most Sundays.)

I encountered a new construction technique for pocketing a few weeks ago. I haven't done any 3-d printing, but one of the other departments at my office 3d-prints internal parts for oceanographic data collection systems. For many of them, they use solid shells for the boundaries, but in the gaps, they trace the printer thread at what looks like about five extrusion diameters apart, and rotate between 60 and 90 degrees between layers, producing an "open foam" pattern which is much lighter but about half as strong as a solid block would be. Bird bones.

Edit:

Originally Posted by Al Skierkiewicz

2016 rules may differ, please be sure to read everything in the robot section, twice!

Twice? Twice? I notice something I missed before every time I read the rules (latest yesterday). I must have read the 2015 rules at least 10 times before bag and tag and probably 30 times so far.

[Cory]

Originally Posted by MrForbes

Mainly, I see the extensive use of CNC material removal as kind of wasteful. This mostly has to do with the fact that I'm cheap and lazy.

If you have the resources to purchase extra material and then cut it away, then go for it!

You make valid points about pocketing. On 254 we feel that lots of teams just pocket away with no understanding of why they're doing it, other than to look cool. Individual teams have different resources. If a team doesn't have much in the way of CNC resources, they probably shouldn't be designing a diamond patterned belly pan...or pocketed gearbox plates.

But the two points above make zero sense to me.

1) If the part is going on a CNC, the time it takes to pocket is trivial in most scenarios. Our gearbox plates would be run maybe 3-4 minutes quicker if they weren't pocketed. The time to setup the machine is the biggest sink, not the run time. I don't see what laziness even has to do with it. It requires more effort to manually mill any kind of gearbox plate than it does to CNC mill it.

2) Your comment about "buying extra material just to throw it away" doesn't make any sense whatsoever. If my gearbox plate fits inside of a 6" long piece of .25" x 6" 6061 bar stock...how am I throwing material out? I'm certainly not pocketing material that is outside of the bounding box of the part. I'm pocketing material that is inside the perimeter of the part no matter what.

I'm glad that you think pocketing isn't necessary for your team, but 9/10 years 254 would not make weight without the heavy use of strategic pocketing.

[GeeTwo]

OK, now I'll take the opposite side:

Originally Posted by Cory

But the two points above make zero sense to me.

1) If the part is going on a CNC, the time it takes to pocket is trivial in most scenarios. Our gearbox plates would be run maybe 3-4 minutes quicker if they weren't pocketed. The time to setup the machine is the biggest sink, not the run time. I don't see what laziness even has to do with it. It requires more effort to manually mill any kind of gearbox plate than it does to CNC mill it.

Got no CNC. Got no manual mill. Got no plasma or laser cutter. Got no router. Got no sheet metal sponsor. Pocketing to us means a drill press or jig saw, or worse. Your 3-4 minutes would take us 3-4 days or more. You're on a different plane of existence from most of FRC.

Originally Posted by Cory

2) Your comment about "buying extra material just to throw it away" doesn't make any sense whatsoever. If my gearbox plate fits inside of a 6" long piece of .25" x 6" 6061 bar stock...how am I throwing material out? I'm certainly not pocketing material that is outside of the bounding box of the part. I'm pocketing material that is inside the perimeter of the part no matter what.

This "makes no sense whatsoever" makes no sense whatsoever. If all goes well, there never was meterial outside of the bounding box of the part. However, every bit of material that was originally inside the bounding box of the part but cut away (i.e. in a pocket) is thrown away (or hopefully recycled, especially if aluminum). I certainly hope you don't discount the material originally in the pockets from your BoM; that would have violated the last several years of rules!

Originally Posted by Cory

I'm glad that you think pocketing isn't necessary for your team, but 9/10 years 254 would not make weight without the heavy use of strategic pocketing.

As I said earlier, you're on a different level of play; birds vs mammals. The key item between pocketing as practiced by top level teams like 254 and most rookie teams is that 254 (please correct me if I'm wrong on this) planned the pocketing from the get-go and knew how much strength would remain, whereas rookie teams pocket out of desperation, without a good understanding of what is acceptable vice what is excessive. For us mid-level teams, it makes much more sense to select the proper strength members than to engineer them by buying over and pocketing down.

[Abhishek R]

Originally Posted by GeeTwo

This "makes no sense whatsoever" makes no sense whatsoever. If all goes well, there never was meterial outside of the bounding box of the part. However, every bit of material that was originally inside the bounding box of the part but cut away (i.e. in a pocket) is thrown away (or hopefully recycled, especially if aluminum). I certainly hope you don't discount the material originally in the pockets from your BoM; that would have violated the last several years of rules!

Say you have a gearbox plate. Whether it was pocketed or not, it will end up on the robot as part of the gearbox. The material that is eventually pocketed was likely not going to be used elsewhere on the robot, so the fact that the material is there or not becomes irrelevant. It then becomes advantageous to just remove it in order to save weight - why not take every advantage you can get, right? I understand it's not much and for most teams there are better things to spend time on and ways to prevent this from becoming a large factor (my team did not pocket most plating on our robots), but if you can do it, might as well.

[Cory]

Originally Posted by GeeTwo

Got no CNC. Got no manual mill. Got no plasma or laser cutter. Got no router. Got no sheet metal sponsor. Pocketing to us means a drill press or jig saw, or worse. Your 3-4 minutes would take us 3-4 days or more. You're on a different plane of existence from most of FRC.

Well you wouldn't be able to make a gearbox plate to pocket, without a mill, so this is kind of an apples and oranges comparison.

Originally Posted by GeeTwo

This "makes no sense whatsoever" makes no sense whatsoever. If all goes well, there never was meterial outside of the bounding box of the part. However, every bit of material that was originally inside the bounding box of the part but cut away (i.e. in a pocket) is thrown away (or hopefully recycled, especially if aluminum). I certainly hope you don't discount the material originally in the pockets from your BoM; that would have violated the last several years of rules!

Unless you only make plates with right angles you're going to have material outside the bounding box of the part. It's a very loaded statement to say "if all goes well, you have no material outside the bounding box". Maybe that's true if your team's goal is to never make a non orthogonal part. That is not the case for many teams in FRC.

You also may consider it semantics, but "contouring" is absolutely different than pocketing and contouring is not under discussion here. Furthermore, teams can and do just bandsaw/belt sand the outer contour of their gearbox to reduce weight or provide clearance for other features of the robot.

Yes, material inside the bounding box of the part that you pocket out is thrown out...but MrForbes is making it sound like it's "extra" material that you didn't have to buy and you could have purchased a smaller piece and saved money. It's metal you already bought. It's metal that cannot be used for anything else. His statement makes absolutely no sense in that context.

Your last point is seemingly irrelevant...unless you're suggesting that you think 254 might be violating the rules. We are well aware of how the BOM works.

[Greg Woelki]

Originally Posted by Cory

Yes, material inside the bounding box of the part that you pocket out is thrown out...but MrForbes is making it sound like it's "extra" material that you didn't have to buy and you could have purchased a smaller piece and saved money. It's metal you already bought. It's metal that cannot be used for anything else. His statement makes absolutely no sense in that context.

I think MrForbes is talking more about using thinner, unpocketed (or minimally pocketed) material when he's talking about paying for material that is "thrown away," since you are clearly right in the context that you considered his comment. It obviously depends on the situation, but I'm sure we can agree that the time and money savings often make thinner material worth considering, especially if whatever part in question is being made on a drill press or manual mill (or even on the very low end CNC mill that my school has).

[GeeTwo]

Originally Posted by Cory

Well you wouldn't be able to make a gearbox plate to pocket, without a mill, so this is kind of an apples and oranges comparison.

More like cherries and pumpkins, but it's good that you understand that the situation is so different as to defy reasonable comparison.

Originally Posted by Cory

Unless you only make plates with right angles you're going to have material outside the bounding box of the part. It's a very loaded statement to say "if all goes well, you have no material outside the bounding box". Maybe that's true if your team's goal is to never make a non orthogonal part. That is not the case for many teams in FRC.

You also may consider it semantics, but "contouring" is absolutely different than pocketing and contouring is not under discussion here. Furthermore, teams can and do just bandsaw/belt sand the outer contour of their gearbox to reduce weight or provide clearance for other features of the robot.

You completely lost me on both of these points. Please elaborate.

Originally Posted by Cory

Yes, material inside the bounding box of the part that you pocket out is thrown out...but MrForbes is making it sound like it's "extra" material that you didn't have to buy and you could have purchased a smaller piece and saved money. It's metal you already bought. It's metal that cannot be used for anything else. His statement makes absolutely no sense in that context.

As I read it, MrForbes' point is that you could have bought thinner material originally, and yes, saved money. The money for thicker stock would never have been spent, and this makes perfect sense in my context.

Originally Posted by Cory

Your last point is seemingly irrelevant...unless you're suggesting that you think 254 might be violating the rules. We are well aware of how the BOM works.

If you were referring to:

Originally Posted by GeeTwo

I certainly hope you don't discount the material originally in the pockets from your BoM; that would have violated the last several years of rules!

I certainly did not mean to imply that you were violating the rules, but rather to express confidence that you were not violating the rules.

[Andrew Schreiber]

Originally Posted by Cory

Well you wouldn't be able to make a gearbox plate to pocket, without a mill, so this is kind of an apples and oranges comparison.

I mean, you don't NEED a mill to make a gearbox. It helps but it's certainly doable on a drill press, I've seen it done. Of course, you could pocket with a drill if you REALLY wanted to as well and just make your bot look like swiss cheese.

The big question is, if you don't have these resources and you don't know how to pocket wouldn't the resources being consumed be better placed by leveraging COTs components during the build season? [1] With the low cost of various planetary boxes and the wide range of reductions available in off the shelf spur gear boxes I question why any team that didn't make Einstein last year would need to design custom boxes [2].


I'll go back to writing code and stop nitpicking.



[1] Obviously not a question directed at you/254, more a general thought.

[2] From a functional standpoint, many of teams have the resources and knowhow so it make sense to use it to save weight/money but for teams that DON'T have resources to do it easily, use COTs.