To rigidify or not to rigidify? - vibration issues
-
I have recently upgraded my delta with new nema 23 motors, new fancy rails, the whole shebang. I am quite happy with the limited testing I did so far but the noise my printer makes has worsened a lot it seems. From what I can tell, it is the actual frame that resonates (at around 800Hz). I can feel the vibrations in the aluminium extrusion towers. In the past I already filled the voids inside the extrusions with sand, but that did not change much.
Short video on YT
Now my question is, do I try to rigidify the frame as much as I can or do I abandon that idea and try to make an sound dampening enclosure?
I was already planning to add some (printed) braces between the towers to serve as a attachment points for my 4th axis. I could fabricate hollow steel tubing in stead of printed pieces, something like the orange parts in the model. Bolted to each of the towers, they could(?) increase rigidity a lot I think/hope. Between and outside of the braces I can still add sound dampening material, or at least that was one of the plans I am playing around with. The main thing that has my worried with this idea is what if I just make the problem worse by introducing more material that can vibrate and resonate.
-
How about a couple strap clamps?
-
I've seen people run fishing line crossbrace members. Maybe that would help with the tension for a cheap proof of concept before committing to more material?
-
What about just adding isolation mounts to the Nema23 motors. They are the source of the vibration so if you isolate them it should be greatly reduced.
-
@mrehorstdmd said in To rigidify or not to rigidify? - vibration issues:
How about a couple strap clamps?
I am reluctant towards putting a permanent preload radially so to say on the towers. I believe the straps could cause the towers to deflect slightly before they become effective. That being said, I will am going to see that straps I have laying around just to try this and see what happens. I can validate that going the rigidify route is the way to go.
@H2B said in To rigidify or not to rigidify? - vibration issues:
I've seen people run fishing line crossbrace members. Maybe that would help with the tension for a cheap proof of concept before committing to more material?
Do you mean something similar to what @mrehorstdmd suggested?
-
@baird1fa said in To rigidify or not to rigidify? - vibration issues:
What about just adding isolation mounts to the Nema23 motors. They are the source of the vibration so if you isolate them it should be greatly reduced.
I already have (homemade) pieces of 2mm NBR between my motor mounts and the steppers, my plan is to try multiple sheets of those stacked and see what happens.
-
Sand won't do much, its not going to be coupled to the aluminium.
Polymer concrete.
Or just mix the sand with some epoxy. People get all worked up about ratios, but basically, sand and epoxy.
Aluminum is really stiff and great at transmitting vibration... adding a bunch of mass (of a type that absorbs vibration, like epoxy filled with aggregate) is the best way.I've been meaning to try adding a metal tube (bolted to the extrusion) filled with water (maybe with a gelling agent), but I've not got to a stage where I actually need it yet. You might want to think about that, as the voids in the extrusions are probably too small a volume.
You'll end up with a very heavy printer... but that's because the only solution to vibration is mass... well, unless you want to get all fancy with active damping, but who knows if that is even a thing in machine/frame design. -
@theruttmeister said in To rigidify or not to rigidify? - vibration issues:
Sand won't do much, its not going to be coupled to the aluminium.
Polymer concrete.
Or just mix the sand with some epoxy. People get all worked up about ratios, but basically, sand and epoxy.
Aluminum is really stiff and great at transmitting vibration... adding a bunch of mass (of a type that absorbs vibration, like epoxy filled with aggregate) is the best way.I've been meaning to try adding a metal tube (bolted to the extrusion) filled with water (maybe with a gelling agent), but I've not got to a stage where I actually need it yet. You might want to think about that, as the voids in the extrusions are probably too small a volume.
You'll end up with a very heavy printer... but that's because the only solution to vibration is mass... well, unless you want to get all fancy with active damping, but who knows if that is even a thing in machine/frame design.Actually, I played around with epoxy granite before in this machine. I cased a heavy machine base from EG for the printer. While it did cross my mind at the time, I did not fill the extrusions with epoxy for two reasons. One, did not have any left over and epoxy ain't cheap
and I was told that the holes in the extrusions would be two small to actually contribute if filled with EG. Do you think figuring out how I am going to get the sand out and filling them back in with EG is the way to go? -
as you don't need any strengt you may get away with any other adhesive that will fully cure and form a reasonable bond to the aluminium?
-
@Nxt-1
I don't think it will add enough mass.I think you need to bolt a big block of epoxy granite to each upright. You could try casting a simple block that you then literally bolt on. Or you could build a mold that lets you cast the epoxy directly onto the aluminum.
I had wondered what that base was made from. That's got to be helping a lot. But you still have 3 very long uprights and 2 big flat plates at the top, which will all resonate like mad.
Heck, it would be interesting to try some off-the-shelf concrete blocks, bolted on. That could probably be done for peanuts. The fancy CNC people tell me that rubber bags filled with concrete is actually the most effective damping option on a metal frame... so maybe concrete blocks with a thin sheet of rubber as a spacer?
-
@theruttmeister said in To rigidify or not to rigidify? - vibration issues:
Heck, it would be interesting to try some off-the-shelf concrete blocks, bolted on. That could probably be done for peanuts. The fancy CNC people tell me that rubber bags filled with concrete is actually the most effective damping option on a metal frame... so maybe concrete blocks with a thin sheet of rubber as a spacer?
The 'bolt a vertical column of concrete' route would be a nice way of adding mass, yet having it relatively easy removable. At a little over 110kg, the machine is already at the max of what I can still move somewhat without involving stuff like rented cranes lol.
I don't really get what you mean with rubber bags filled with concrete, I fail to imagine a rubber bag. Do you perhaps have a picture?
-
@Nxt-1
I do not.
But imagine you made your printer (or CNC mill) from box section steel. You then stuff a rubber bag (like an inner tube) down the inside of the box section. Then you pump the rubber bag full of concrete, till the bag has stretched and the box section is now full of concrete wrapped in rubber. Then let it cure.
Presto, one very strong, very cheap, mostly concrete beam. Which they tell me is a very effective at vibration dampening. -
@theruttmeister I see, I doubt I am going to be able to pull that off actually
I wonder if I could buy of the shelf concrete poles, like for fencing use and drill holes trough them. Then with a thin piece of rubber bolt them to the outside of the frame, as you suggested. The ones in the picture are 10x10cm and as long as I would ever need.
-
I hope you have a good strong floor to support this monstrosity.
-
@Phaedrux said in To rigidify or not to rigidify? - vibration issues:
I hope you have a good strong floor to support this monstrosity.
From what I find on forums in my country, load limits for a 1st floor vary wildly from 200 to 2000 kg/m². Most people seem to agree on ~300kg/m² though. I would be looking towards those values but should stay below them by a good percentage. For example, the current printer + 3 of the above poles would be ~200 kg
-
From my time with fish tanks I recall that it's usually safer closer to the wall than the center of the room.
-
@Nxt-1
Perhaps a little bigger than I was imagining... but I don't see why they wouldn't work. Plus I bet they are cheap!Location and number of the bolts might be a variable.
And the rubber would help reduce the risk of deforming the extrusions to match the no doubt not flat or true concrete. -
@theruttmeister Since sleeping over it I am thinking about casting 7.5x15x145cm pillars myself (the extrusions are 15cm wide and 150cm high). It would cost me about €20 in concrete mix for all pillars instead of €40 a piece for of the shelf pillars (which I would need to cut to length as well)
I do wonder if my current approach of using M5 bolts and t-nut inserts will suffice to attach the concrete slabs to the towers. If not things start the get nasty real quick, as I'd need to drill larger holes straight through the extrusions
@Nxt-1 said in To rigidify or not to rigidify? - vibration issues:
@baird1fa said in To rigidify or not to rigidify? - vibration issues:
What about just adding isolation mounts to the Nema23 motors. They are the source of the vibration so if you isolate them it should be greatly reduced.
I already have (homemade) pieces of 2mm NBR between my motor mounts and the steppers, my plan is to try multiple sheets of those stacked and see what happens.
I just installed a total of 10mm of these NBR sheets and it improved the coupling a bit but not enough. I think a lot of the noise gets coupled via the bearings that support the shaft in the motor bracket I designed.
-
@Nxt-1 The pics I have seen were of adding fishing line cross-bracing the upright frame members to act as triangular support segments so that the frame itself becomes more rigid. I don't have a picture on hand, but I'm sure if you google around you may find some implementations of that.
-
@H2B
I wonder how applicable that technique is to the size of machine I have.@mrehorstdmd said in To rigidify or not to rigidify? - vibration issues:
How about a couple strap clamps?
I tried with the only strap clamp I could find and did not really notice a significant difference sadly, worth the try though.
@theruttmeister
I got myself four bag of ready to concrete mix and will play with it, hopefully in the coming days. Most likely I will go with the M5 bolt+t-slot nut approach as drilling holes trough the extrusions is just asking for troubles with all the interrupted cuts, at least with the tool I have at my disposal. I'll report my findings once I have something to report.
