UniStrut Frame designs

hey yinz, and ya’ll,

This thread is a splinter of the Frame Design Thread, detailing a frame build using UniStrut (a brandname, many competing and similar materials are of equal capability).

Big thanks to @bdillahu for pioneering this frame design here:

Unistrut is a metal framing system primarily intended for hanging pipes and conduit and more or less readily available at any Home Improvement Store.

This frame is inherently strong and less susceptible to flexing than the wood frame when assembled, and has the added benefit of completely disassembling and reassembling without wear on major parts. This frame can be made to collapse/fold/partially disassemble much more easily than the default wood frame. (although we have no documented folding models just yet)

Note that UniStrut is more expensive than wood and screws, particularly in the attaching hardware. Those bits add up quite quickly too, with an expected full frame design costing somewhere in the range of ~$250 all told.

More details to follow, but take a look at some of @bdillahu 's frame pictures and original discussion here:

cheers

6 Likes

Here are some annotated pictures of my UniStrut Frame…

1st, the mostly finished product: (Note the lack of bricks/weights, chain tensioners, casters, etc, theres still some work to get it tip top)

Aside from some additional full length (10’) cross members, this design is identical to @bdillahu 's original design as far as the cut UniStrut is concerned.

Note: due to the measly selection at my local home center stores, I could only find 1/2" hole L-Brackets, and exclusively used 3/8" bolts and ‘strut-nuts’ (specially shaped to grip the overturned channels). This actually worked really well as the 1/8" all around wobble allowed enough give to set everything in the correct position more easily than a tight fit of a 3/8" hole on L-brackets would have. It’s counterintuitive but definitely helped along the way.

Starting at the back of the left side (as facing the machine) is a simple L-bracket, long side up:

Moving to the front of the bottom strut, you’ll see that 2x L-Brackets are set with one Long side up on the floor strut, one short side out on the vertical strut:

Returning to the back of the frame, we see the 2x L-bracket ‘T’, This longitudinal connection uses a 2" long 3/8" bolt, washers on both sides and a standard nut. The ‘T’ connection greatly assists in getting the frame square and level while assembling, which is a little tricky without them as everything wants to wobble and twist. Once the ‘T’ connections are in place on the horizontal members you can easily slide the vertical elements around with one element of the ‘T’ connection loose, and the other tightened and aligned:

Next is the connection of the rear vertical strut forming the pointy end of the triangle, along with the front facing horizontal support:

Note how the horizontal strut is both set back (on the long L-bracket, in a ‘T’) and the rear strut is made to rest on the front horizontal strut:

Once you’re this far along, you’ll have a frame thats standing up and sturdy enough to really dial in your connections, make sure its all level, and there’s no twist and tighten it up to give it some strength:
IMG_2357

Next is the assembly to hold the Motor mounts. Another ‘T’ connection on the 7" top bars, and the only non-standard connection, 3 large washers used to span the gap. Make sure these washers are big enough (1.5") to securely grip the channel slides, or they will bend/slide into the channel when tightened up.:

The top most full height cross member is heavy enough to conk you on the head if you’re not careful, and it likes to slide until its tight, so you’ve been warned…

This is where we move onto the MaslowCNC motor mounts these bolts are 1/4" wide and might be 2.5" long, I can’t recall, but were knocking around the garage. They’re applied with generous washers:

And thats about it.

I opted to keep the ‘work piece crossbar’ plain metal for now, and still have some alignment work to do to set up a spoilboard. The frame does NOT need a sheet of plywood in place to maintain its stiffness and I’m thinking that a 1.25" rigid foamboard as a spoilboard might leave me with just enough of a lip (3/8") to balance many different thickness of stock without much further modification.

Once I get the machine dialed in and running I have parts for joining 2 struts lengthwise, and will be investigating the maximum width a stock maslow can be made w/o additional chain.

Also, it appears that I can use the motor’s own screw mounts to install the chain rollers, it’ll be a close fit but I’m pretty sure they’ll work well and keep the chain well attached without interfering with the outrunning chain.

anyway, good fun building this, I cant wait to get the last bits in place and see what she can do!

cheers,

15 Likes

Wow, that’s great looking… I’m going to steal your “double right angle” joint idea for the bottom front angle I think. Part of why my bottom leg is on edge is I didn’t come up with the (now obvious) way to attach it at a non-90deg joint :slight_smile:

I used a sheet of ply bolted to the frame for stiffness and you used the additional unistrut… yours certainly looks like a good way. I have been using a spoil board over the “part of the frame” ply, then the workpiece.

One question - just curious - why did you mount your motors “in front” of the angle bracket? Any particular reason? I mounted mine behind it with the shaft through the hole and it seems to work.

Your way does give clearance for the bolt - I’m betting that’s the reason :slight_smile:

1 Like

hey, thanks!

for sure, steal that joint. I had to come up with a way that didn’t leave a twist or needing a big pile of washers and a new one-off sized bolt to make up space… I think keeping 2 of the 3 sides of the triangle in the same plane was a good idea. I had it level and tight already, but you can slide the vertical down to stop on the bottom strut while still keeping the lean angle wherever you want, so thats a nice benefit i think…

As for the motor mount, the bolts to the strut were not an issue, even with a washer it cleared by just enough.

I’ve set the motor mounts up both ways while i was building it, and found that with 2 new screws long enough for the chain roller plastic, I can probably mount the chain rollers right in the lower mount holes if i mount he motors to the back. I’ll post more pictures once I have that together.

I’ll definitely recommend getting rid of the need for plywood for rigidity.

Its immediately obvious that with some hinges and flat brackets the whole thing could be made to assemble and disassemble down to the frame triangles and a small stack of folded cross members in 10-20mins… easily down to ~1/3 the size of a ‘wall folder’ type frame design speculated on…

Great for keeping the garage tidy!

The only thing preventing me from going that route from the start is that the home center stores I visited (4!) all had a minimal selection of accessory parts. They’ll need to be special ordered at some point once the garage gets re-cluttered.

Anyway, thanks for the inspiration, your work was a huge help getting me on the right path!.

cheers

1 Like

very nice. As you are looking to mount the idler roller, remember that the motors can be rotated in any direction if the hole isn’t quite right and the shaft is off-center.

That’s making me question my wood one :slight_smile:
I may build the wood one as a reference, and then a metal one to use more commonly

@mrfugu I know you spent quite a bit of time writing up your unistrut build, but since I dont want to buy the wrong materials is there any chance of you putting together a materials list? This has been tops on my priority list, to “solidify” the frame. With the flex from the standard frame I feel this will expand our box of accuracy!!
Thanks for all that you do!!!

yeah, I will eventually, I haven’t yet got my machine fully online, and my schedule has some big travel gaps coming up, so it realistically won’t be before the new year until i can distill everything into a Wiki page.

here’s a cheat list off the top of my head, the numbers might give you some extra, or they might not, but you’ll get close enough

6x “1/2 height” 10’ Superstrut (Home depot/lowes brand unistrut)
1x “full height” 10’ Superstrut

at least:
24x L-Brackets
48x 3/8"x1" bolts
48x 3/8" strut-nuts
10x 3/8"x2" bolts
10x 3/8" standard (or airplane) nuts
1 box 3/8"x1" steel washers
12x 3/8"x2" steel washers
4x 1/4"x2.5" bolts
12x 1/4"x1.5" steel washers
4x 1/4" standard (or airplane) nuts

I think thats more or less it. I have a few spare L-Brackets, and bought the nuts and bolts in boxes of ~20-25/ea i think, and the strut-nuts come packaged x5, so there’s spares there, but not so much i feel I wasted much $$, maybe $10-20 in spares left over, but I’m planning on adding some length to the motor bar, and figured the design will change a bit. Spares are always good, and save time in the end…

good luck!

6 Likes

Perfect, Thanks @mrfugu !!!

you know what, it might be 6x thin unistruts… 3 for the 2 ‘triangle uprights’, and 3 horizontal.

I updated the build list above.

1 Like

those 1/4 bolts might be 3" long as well. I’m not there to measure atm.

good luck!

1 Like

Keep in mind that there is really nothing that’s dimension critical here.

See the “musings on frame design thread”,

This is a ‘top beam’ design, using a 10’ beam (because that’s the size the material comes in, and it’s about the size of the stock maslow)

This beam needs to be positioned out over the work area (ideally in a way that allows adjustment of how far out it is)

The rest of it is just a way to hold the beam and workpiece in place, and can be just about anything.

Exellent build-log @mrfugu !

@dlang, for sure, I understand that its all arbitrary to an individual machine’s measurements, but lengthening (and/or heightening) the beam the motors are on increases accuracy generally.

I’ve avoided the physics really until now… but:

At some point (a theoretically calculated distance) from any maslow using spec chain weight and spec motor strengths, chain sag will increase with the amount of chain deployed to a place where accuracy suffers and cannot be easily compensated as the free motion available via sag is too high.

There is also a hard limit of some (unknown theoretical angle) relative of near side motor to the far side of the workpiece, beyond which the motors nor software can compensate (assuming they will be able to in the future) for sag.

Assuming a stiff metal frame, I can increase length (acknowledging the diminishing returns of the calculated limits above) more easily than height (both for available room in my garage, and the aforementioned variables)

Assuming the math for (some amount of compensation for) sag is developed eventually the machine is more likely to be ‘ideally’ (for stock chain/motor hardware) wider than taller.

I grabbed an extra 10’ strut with the idea being that once things mature a bit and I have as accurate a machine i can make that is 10’ wide, i can stretch out another little bit (estimating ~12-13’ total width) and hopefully increase that ‘highly accurate’ area a little wider.

After we’ve added all that can be added code wise to compensate for chain sag, we’ll be able to calculate the ‘optimal’ height and width that the stock chain/motor combo can handle for maximum size accurate work area.

i mean, i think so…

2 Likes

@dlang, for sure, I understand that its all arbitrary to an individual
machine’s measurements, but lengthening (and/or heightening) the beam the
motors are on increases accuracy generally.

yes.

I’ve avoided the physics really until now… but:

At some point (a theoretically calculated distance) from any maslow using spec
chain weight and spec motor strengths, chain sag will increase with the amount
of chain deployed to a place where accuracy suffers and cannot be easily
compensated as the free motion available via sag is too high.

more precisely, the force of gravity that results in tension on the chain
becomes low enough that the sled may stick instead of moving.

There is also a hard limit of some (unknown theoretical angle) relative of
near side motor to the far side of the workpiece, beyond which the motors nor
software can compensate (assuming they will be able to in the future) for sag.

vertical :slight_smile: other than that, it’s not a hard limit, it’s the sled sticking.

Assuming a stiff metal frame, I can increase length (acknowledging the
diminishing returns of the calculated limits above) more easily than height
(both for available room in my garage, and the aforementioned variables)

the only part that needs to be particularly stiff is the top beam, the rest
isn’t under heavy forces.

Assuming the math for (some amount of compensation for) sag is developed
eventually the machine is more likely to be ‘ideally’ (for stock chain/motor
hardware) wider than taller.

since the stock we are working on is much wider than tall, that is a safe bet
:wink:

I grabbed an extra 10’ strut with the idea being that once things mature a bit
and I have as accurate a machine i can make that is 10’ wide, i can stretch
out another little bit (estimating ~12-13’ total width) and hopefully increase
that ‘highly accurate’ area a little wider.

worth experimenting with

After we’ve added all that can be added code wise to compensate for chain sag,
we’ll be able to calculate the ‘optimal’ height and width that the stock
chain/motor combo can handle for maximum size accurate work area.

I expect it’s always going to be a compromise on space, but yes.

right on. yeah there’s sled weight too obviously…

but my main observation/guess/assumption is that somewhere within ~12-14’ width will be a sweet spot that increases the ‘highest measurable accuracy’ window to, or beyond, a ‘full sheet work area’ and that that machine won’t be unreasonably tall (< 8’) and (if using a unistrut construction) easily assembled/disassembled.

If that’s the case, I would call it good for a v2.0 design. wouldn’t you?

Wow, this is a really nice design. Structurally unistrut is hard to beat for rigidity. Thanks for the inspiration.

Do you think that a full thickness unistrut for the sheet support would be an improvement in any way? (lowering the center of gravity, additional rigidity, etc)

I see you put sand bags on the frame to dampen vibration and presumably keep the frame from walking across the garage.

thanks, but the inspiration def goes to @bdillahu.

I don’t think full height is needed on the workpiece support. the weight of the sheet is spread evenly enough that its not an issue. Those sand bags were temporary to allow me to true everything up more easily, and probably aren’t needed in a complete frame, not that it hurts…

good luck!

2 Likes

Thanks, but you’ve certainly taken things to the next level!

1 Like

Is 12ft or 13ft really the sweet spot for the motors? If so, I will use a 10ft span for the top and then cut 2 shorter 3ft sections and bolt them to the front of the 10ft with 1.5ft overlap per side my motors will be 2-3 inches up and I’ve added 3ft total.

I could use large wing nuts on these bolts, do allow me to quickly swing these two arms in to get back to 10ft wide overall. I think I want to imagine a wall mount version and a cart version of unistrut. I also plan to make a sled that can let a miter saw move left-right 4 ft to facilitate quick chop operations (like Home Depot does with their plywood saw)

Is 12ft or 13ft really the sweet spot for the motors?

we don’t know the sweet spot for sure. We know that as the chains get closer to
vertical you have more trouble, but we don’ thave enough people who have built
and reported on wider frams to give a good comparison (let alone anyone who has
built a standard width and a wider frame to give us a direct comparison)

If so, I will use a 10ft span for the top and then cut 2 shorter 3ft sections
and bolt them to the front of the 10ft with 1.5ft overlap per side my motors
will be 2-3 inches up and I’ve added 3ft total.

I would start with just a 10’ wide frame, get it working, and then have the arms
optional so you can do a direct head-to-head comparison and see how much it
helps (the biggest difference would be in the bottom corners, especially at high
speeds)

I could use large wing nuts on these bolts, do allow me to quickly swing these
two arms in to get back to 10ft wide overall. I think I want to imagine a
wall mount version and a cart version of unistrut.

That sounds good

I also plan to make a sled that can let a miter saw move left-right 4 ft to
facilitate quick chop operations (like Home Depot does with their plywood saw)

The problem I would see with this is that the sled can rotate, so as you move
the saw off center, the sled will rotate due to the weight.