This fall, 3005 was given an additional classroom with enough space to set up a full FRC field.
In the past we have practiced on one half of our school’s practice gym which has been a huge resource for us. There was enough space to set up a full field but the space would also get used for events like job fairs a few times throughout the season. We also weren’t able to have it set up outside of the build competition season so FRC offseason development has been largely nonexistent for us. We kept the number of field elements, especially the perimeter, to a minimum to facilitate easy setup and teardown.
We’ve always felt the need for a higher fidelity field perimeter. With the new space, we have the ability to commit to a more permanent setup. The AM perimeter (or fabricating our own metal & plastic perimeter) is out of reach for us for the foreseeable future so we set out to design our own perimeter.
Construct the perimeter with low-cost materials (wood) and low-labor methods (CNC router).
Achieve a robust perimeter to keep robots contained and protect bystanders.
Feature a full perimeter (when game-specifics are included) to keep gamepieces contained.
Follow the same segmenting and attachment pin locations as the real field.
Realistic alliance station experience for sightlines, driver positioning, and lack of auditory feedback from the robot.
Realistic field access (gates) for realistic match setup practice.
A few highlights/details:
The elements are constructed with a notch-and-tab build style out of 1/2in plywood. There are also a few 2x4s. The pieces are fastened using a gap-filling construction adhesive and held in place with small nails or wood staples. The rest of the parts are COTS.
The elements are designed to have a large surface area in contact with the floor and use hook tape to transfer most of the load into the carpet. We DON’T expect the bolts connecting the field segments together to be the primary load path during impacts.
Number 4 above is a key point of this perimeter system. This is especially important for us when testing vision systems and full-field odometry. Setting up a practice field to have realistic field element positions can be very difficult when the attachment method for the practice elements is different than the real field. The 3005 perimeter uses the same corner pin method that the real fields use to form angled corners and similar field features. This allows our field to be set up using the exact same chalk line and measuring steps that the real fields use and will make it much easier to get a dimensionally realistic field setup.
We are still in the process of building this but wanted to share the project with the community before kickoff. I will report back with photos once more elements are built.
Here is a link to the CAD:
We will either design our own game-specific field elements right after kickoff or adapt existing CNC-routable designs to fit this perimeter. Whatever field elements we use will be made publicly available and we’ll link them in this thread. We will also make another assembly in the Onshape document for whatever the 2025 perimeter arrangement is.
This is beautiful, really nicely thought out design - particularly the extra Velcro surface in several areas.
I’d be very concerned about the perimeter panels actually being strong enough to contain a full energy robot interaction. Having watched 1/2" polycarbonate get shattered twice in the offseason, it’s hard to believe 1/2" ply will hold under full energy impact.
(We’ve tipped over our lab tables (1"-ish thick tabletop) for a sufficient perimeter in the past, and I’m planning to use stacked 2x4 in the 973/581 style in the future.)
Yeah 1/2" ply isn’t going to hold up to real use. It might if lined with Polycarb, but without a frame rail behind it still likely can blow sections out.
Very cool idea, but I worry about strength. We pocked a lot of holes in our old plywood field, and this construction doesn’t offer an easy way to make repairs.
If you’re using a CNC router, the tab and slot system will be way faster to assemble and more accurate too as the pieces will be self locating. Probably cheaper because of fewer parts required, though I’m not sure what the cost comparison is between 2x4s vs. construction adhesive tubes.
@Orion.DeYoe - I’d definitely add some 2" holes in the lower corners of all the parts so you can run power and/or Ethernet out of the way all the way around the perimeter in the future if you eventually set up something like Cheezy Area. Looks very cool!
I can tell you, having built some cnc-routed plywood field elements, that it is very not faster than cutting panels and 2x4s with a circular saw or table saw.
If the field is wood, it will get busted up, but fear not. Typically ,the field is still useable (i.e. in the pic that @nuclearnerd posted, the field is damaged but can still be used while someone figures out how to fix it. Don’t fall behind on fixing/maintaining it or before you know it, you have a pile of kindling).
The biggest problem we struggled with, especially in the Age of April Tags, is field movement. If you’re big on April Tags and they’re attached to field walls then field calibration becomes difficult. I’m not sure how to address this short of bolting stuff to the floor, which for most, isn’t an option.
The big thing that drives the cost is the quality/source of the plywood. We spent about $3k on the wood but that includes some extras for building game specifics so we don’t have to immediately go to the store on kickoff day.
You need about 40 sheets for the full set of elements.
Our hardware order was ~$500 from McMaster. There are also some other odds and ends you’ll need (nails, glue, hook tape, etc.).
Safe to say you can build a full set of perimeter elements for <$4k.
To address the strength concerns, I have no expectation that this field perimeter will survive the kind of abuse that the real field does (or doesn’t). The design choices for this were made to balance cost, manufacturability, and strength.
Based on the survival rate of several of our high-impact practice elements from the past, I’m optimistic about the strength of this. There are stiffening ribs in the bumper zone on the elements with long unsupported spans which should help.
That being said, I will definitely report back during/after the season on how this holds up and make design changes as needed.
Generally, in our experience, the notch and tab routed construction method is quite a bit faster. Even if they’re about even, it’s a skill we like to exercise because we use it for a lot of robot parts and other stuff around the shop. The other thing is that the tolerances of the parts are decoupled from the skill of the machine operator which is not the case with saw cutting and assembling lumber.
In terms of cost, we want to skin the inside surface of the field so we’d end up using a large amount of plywood anyways. I can’t say for sure that it wouldn’t be cheaper overall but my gut feel is it wouldn’t be a significant improvement.
That’s a great idea! We are definitely going to run power to each alliance station for laptops so this will be a feature we’ll wish we’d added originally.
I think you can retrofit 2x4s in place here pretty cheaply and easily. Just cut something a bit shorter than the gap from angle support to angle support. Do 2-5 per span and you’ll be pretty happy with the extra robustness.
Based on our experiences, I’d also recommend lining the field side with Polycarb. This better emulates the bumper to border interaction of the real field, and also protects the wood.
Lets talk about strength. In my experience with non-andymark official fields, fields see the most damage to driver station walls and season specific elements, damaged side walls are pretty rare because they don’t get hit nearly as hard. The official driver stations are an aluminum plate welded to the front of a tube. This structure of this plywood field is basically a U channel with unequal legs.
CAD can easily give us an area moment of inertia for these two shapes, and that comes to 2in^4 for the aluminum and 1870in^4 for the wood. Stress from an impact is proportional to Load * distance from the neutral plane (y) / area moment of inertia (I). y/I wood = ~.0064, y/I aluminum = ~.3 a difference of about 45x. So for the fields to be of equal strength the aluminum needs to have 45x higher yield strength. 6061-T6 aluminum has a yield strength of 40,000psi, plywood strength is hard to estimate but woods used to make it range from 8-18,000 PSI. I skipped over a lot of the details here and made some modeling assumptions, this should really be done with FEA and proper estimated loads if we wanted accurate results but this suggests that 3005’s plywood driver stations are several times stronger that official driver stations.
I would cover the bottom foot or so with 1/16" abs to avoid splinters in the bumpers and provide some protection to the wood from scrapes and dents.
This is a great design, and I would be copying it if we were building a field right now.
We have built a similar field perimeter. I would definitely suggest getting some 10lbs or better sandbags to put on those velcro outriggers. We have seen them get pushed out (rather easily) and the weight can help.
Interesting, I field supervised the FIM Muskegon district last year and 3 of our 6 driver station walls had significant enough bends that they were taken for repairs before the next competition, I believe that 3 more driver stations were identified at other FIM events for repair that weekend. The only side wall damage I saw were 3 sheared rivet (2 of which were noticed during setup). I didn’t see side wall poly brake at any of the 5 events I attended this year where I would have expected to hear about it.
That looks great. It should give your team aneven greater advantage. It is also nice to see yet another full field in Texas.
Something to watch out for is damage to the carpet, so budget for replacing it every few years.
The administrator of the STEM Center in Katy has scheduled replacement of the carpet after just one year. With the driver station walls firmly attached to the carpet, robots from several of the 7 teams based there crashing into the field at near full speed to practice scoring has bowed the driver station walls out and stretched the carpet. The carpet eventually tore across the field, approximately around the Center Line. If your team is the only one practicing for most of the time, this is less likely to happen.
Hope that my team can meet up with yours in April.
Just to be clear, the two of you are talking about two different types of perimeters. Emphasis added:
The AndyMark fields have a different gate design, so particularly if you’re seeing damages to the polycarb near the gates, that could be one factor.
Not sure if this affects your analysis, but you likely know that there are also some “feet” installed on hinges that provide bracing for this tube. The Bri-Weld (non-AndyMark) fields, at least in FIM, have two permanently installed, and we were adding a third removable foot at some events this year. (Not sure if the extras had been shipped out in time for Muskegon.) I’m not sure which axis/axes your moment of inertia is referring to (or really how to describe it appropriately in text form), but it looks to me like your diagrams are from the perspective of a side wall looking towards the other side wall - so maybe you’re referring to impacts that leave a “C”-shaped dent visible from that perspective? I believe at least some of the damaged alliance panels we saw this year left C-shaped dents visible from the top down, i.e. from impacts in the bottom middle of the panel, where the built-in feet and “P” supports don’t provide as much support. I could be misinterpreting, and I still agree that the additional brace in 3005’s design helps in any case.
Yes the bent driver stations are c shaped from above I don’t remember if we had started adding the extra Velcro plates by Muskegon but I didn’t consider the reinforcement from the or the Ps because those are not weak points. Carpet is really quite stretchy at the forces involved in robot collisions though so I don’t think more Velcro plates are going to do much for protecting the walls.
