We have an older Markforged Onyx One that has been super reliable and a less old Onyx Pro we got at the VERY tail end of the broad FRC sponsorship deals on printers Markforged had.
We have not used the Onyx Pro’s continuous fiberglass capability much at all because Onyx alone has been good for most things, and if we need something stronger, then we usually make it out of aluminum (or maybe polycarb).
Curious: What are the killer applications you’ve seen in FRC for the Onyx Pro’s inlaid fiberglass capability? If there aren’t any, I’m considering recommending the school sell it since it is low hours & has considerable value & buy some latest gen fast printers.
Let me open by saying that I don’t believe continuous fiber is NEEDED. The machines and the material are expensive and then there’s the maintenance etc. There is always a way around it, i.e. design it out using aluminum. There is also a valid argument that fiber is overused “This part needs to be strong. Let’s put fiber in it just because!” We print a lot of our robot and find only a few instances that we feel the need.
Here are some examples of where we have used continuous fiber:
These hold the wheels and are quite abused. Carbon fiber (verses Kevlar) to stiffen them. Flexing causes changes in the bevel gear backlash. These are super stiff!
There were 4 of these. This held the elevator to the chassis (square hole) was the hinge point to tilt the elevator (round hole) and held a Falcon while keeping a chain tight. There was a bunch of twisting, torsional, cantilever action going on here. Kevlar was chosen because it’s forgiving.
Work around: Don’t design such a silly thing!
These are the only parts in our 2023 robot that had continuous fiber. Most of the work arounds are straight forward.
All this being said, I lean towards selling for more faster printers unless the “Coolness Factor Bug” bites you then, well, you have to keep it!
Fiber reinforcement gives you much stiffer and stronger parts compared to other 3D printing options (even solid onyx). There are definitely workaround for some parts, but if you don’t have the time to redesign or machine something it’s a great option. I highly recommend reading the design guide to get the most out of it.
We don’t have many machining resources on our team, but have a two mentors from Markforged (including myself). We used our team’s Onyx One for the main pieces connecting the chassis to the upper frame. They held up all season through several hard impacts that bent our frame and a few times the robot tipped over.
Need meaning the part would break without continuous fiber or need in that using continuous fiber provides any benefit? Even for parts that won’t break, running continuous fibers for the extra stiffness is a completely valid use that translates into actual benefits.
Though it is a bit different (we use a Mark Two with Kevlar inlay) we have found it to be perfect for structural parts (mounting gussets, etc) that woud be difficult or time-consuming to precisely machine out of aluminum in house (particularly parts that woud require a 5th axis). In fact nearly every bit of mounting hardware on our most recent bot (including the chain attachment point for our elevator) was printed in Onyx with kevlar inlay. All without a single failure. Additionally we have, for years, manufactured all of our pullies out of Onyx without CFR.
Could you go into a bit more depth on how you choose what fiber to reinforce with? Weve had botlh fiberglass and CF sitting around for years now but someone ages ago seems to have decided kevlar was best and we’ve simply stuck with it for all applications.
Without continuous fiber, the saddle would “parallelogram” when turning hard and the bevel gears would either be crammed together or spread apart. When spread apart, the gears would jump and skip. The saddle didn’t break but it is a definite operation failure. Can’t be jumpin’ gears.
Azimuth pulley:
Without continuous fiber, the pulley would eventually start slipping around the hex. This only shows up if the azimuth encoder is on the motor side vs. on the module side. If on the module side, the encoder doesn’t see the slip but the more the slip happens, the weaker the pulley gets, and azimuth performance is degraded and eventually it stops working all together. Pulley didn’t actually break, but the module stops working.
Elevator Frame Mount:
This one wasn’t tested but without continuous fiber we feel that we needed hoop strength on the square hole because the elevator was putting a lot of twisting action on that, so we feared the same phenomena that happens to the azimuth hex interface. We also feared that the Falcon mount would bend or break from getting pulled hard by elevator drive chains.
With us, in the beginning, we would add CF with no serious thought on if it was needed. The common statement was “It makes it stronger”. Sure, it’s stronger but did we get any benefit from it? Many times, it just made the part more expensive. Our attitude now is “Its a tension device, a crack stopper and a stiffener”.
We use either CF or Kevlar (not the fiberglass). In a nutshell, CF is used for rigidity or stiffness and Kevlar is used where flexing will happen or desired. In the examples I listed, chances are, either CF of Kevlar could be used but the goal for the saddle and the pulley is not to flex ever! With the elevator mount the goal was a little flex is desired, part needs to be tough.
Here’s an example of where we needed LOTS of flex:
This is an intake link that would bend like 1/8" polycarb and had Kevlar in it. If we put CF in there, once it gets bent, the CF would snap inside the plastic, and its advantage goes away. In practice, when this part would break, the PLASTIC would crack, and the Kevlar kept it together and the part still worked.
