| Author |
Message |
    
Gerald_D
Registered
Username: Gerald_d
Post Number: 15
Registered: 11-2005
| | Posted on Thursday, December 15, 2005 - 09:46 pm: | 
|
The 8mm screws holding the rails to the gantry are tapped into "spiked" holes
 |
Gerald D
Unregistered
| | Posted on Sunday, December 11, 2005 - 06:15 am: |
|
Tack welds added to retain parallel rails during squaring
Can also see the tacks to the unistrut at each bolt. |
Gerald_D
Registered
Username: Gerald_d
Post Number: 274
Registered: 11-2005
| | Posted on Monday, October 09, 2006 - 06:43 pm: |
|
With some ShopBotters now contemplating some drastic mods to their machines, its maybe time to revive this thread...
(Also see this thread at the SB forum)
The standard PRT gantry consists of 2 pieces of 3"x1.5"x3/16" angle iron rail stiffened by 2.5"x1.625" utility strut. I think it is a given that folk want to retain the angle iron rails, but that they want to do the type of thing that Steve did right at the top of this thread: replace the utility strut and gusset the corners.
Steve retained the functionality of the "hold-down" rollers, whereas I believe they may be doing more harm than good.....
The idea here is to revive this topic and I want to see if I can produce some Americanised (ASTM etc.) plans in inches that you guys could take to local welding shops out there. Any takers?  |
Gerald_D
Registered
Username: Gerald_d
Post Number: 275
Registered: 11-2005
| | Posted on Monday, October 09, 2006 - 07:01 pm: |
|
Our PRT gantry from year 2000:
The "Unistrut" used by SB is type P5500 which is 12 gauge steel, 1 5/8" wide X 2 7/16" deep. For carrying weight the important parameter is the "I-modulus", and for this particular unistrut the I-value 0.523 in4
For resisting "twist" (caused by the wheel load being way off center), the support tube should be closed as a "box-section". Open "channels", like the unistrut, have a very low resistance to twist - so low that data is hard to find. (The rail on top does not "close" the channel unless it is fully welded along the whole length). My estimate of the J-modulus is about 0.003 in4. |
Gerald_D
Registered
Username: Gerald_d
Post Number: 276
Registered: 11-2005
| | Posted on Monday, October 09, 2006 - 07:21 pm: |
|
Compared to the first 4foot MechMate gantry that was retrofitted to the standard 48x96PRT:
This gantry was built "heavy" to see if the extra mass was detrimental to performance. It wasn't. The next 6foot gantry was of 0.08" thick material for the big MechMate. That also works fine.
Here is a table from Allied Tube.....
- 4"x2" 14gage tube: 3.3lbs/ft I=1.02in4
if a 4" high tube is going to reduce your gantry clearance, then
- 3"x2" 10gage tube: 4.3lbs/ft I=1.04in4
The two examples above are to illustrate what paramaters can be played with.
However, aside from the torsional stiffness, the other major shortcoming of the Unistrut gantry is the way in which the struts are (not)connected to the wheel carriers at the ends........ |
Gerald_D
Registered
Username: Gerald_d
Post Number: 279
Registered: 11-2005
| | Posted on Tuesday, October 10, 2006 - 12:19 pm: |
|
This is the PRT with Unistrut:
The biggest flaw is the gap between points A and B. The unistrut is the main carrier of the rail, but it relies on the relatively flimsy rail to connect to the wheel point.
This is how a longer 3"x2" rectangular tube can do the job:
Now have another look at Steve's photos right at the top of this thread. Makes sense now, doesn't it?
Apart from making an overall stiffer gantry, (and not really heavier because those square tube spacers come out), the other huge advantage is that there is now space to beef up the y-car. That redundant square is the biggest obstacle to making a proper y-car....which will be in a thread of its own. (link)
(A word of caution, if you do make the gantry stiffer, then your X-rails need to be precisely parallel. With the original "flexible" design, the gantry ends can flex to accommodate small deviations of the rails) |
Gerald_D
Registered
Username: Gerald_d
Post Number: 280
Registered: 11-2005
| | Posted on Tuesday, October 10, 2006 - 12:41 pm: |
|
Here is a zipped .dxf file of my rough sketchings (to scale) if anybody wants to run with some ideas:
|
Mike John
Registered
Username: Mikejohn
Post Number: 85
Registered: 12-2005
| | Posted on Tuesday, October 10, 2006 - 04:24 pm: |
|
Gerald
My PRT strut, which I believe to be similar to the Alpha, is some sort of aluminium, approx 72mm high, 40mm wide.
Do I need to think about strengthening?
.........Mike |
Gerald_D
Registered
Username: Gerald_d
Post Number: 283
Registered: 11-2005
| | Posted on Tuesday, October 10, 2006 - 04:56 pm: |
|
Mike, As far as I can recall, you have never had "chatter" issues, or anything else possibly related to a flexing gantry. Best to leave it alone.
I am not convinced that the alu is any stronger than the unistrut - the apparent improvement in the "Alpha" gantry seems entirely due to the fact that A is now connected to B. The guys with unistrut gantries could easily weld a patch across that gap as well.
Which leads to the point that the gantries are easy to fix, but that the y-car needs lots of attention which isn't just a patch. Ideally, the y-car needs to be replaced.....we can talk about that in the other thread. |
Robert Cheal
Registered
Username: Robert_cheal
Post Number: 10
Registered: 08-2006
| | Posted on Wednesday, October 11, 2006 - 05:11 am: |
|
Gerald,
The idea here is to revive this topic and I want to see if I can produce some Americanised (ASTM etc.) plans in inches that you guys could take to local welding shops out there. Any takers? |
Robert Cheal
Registered
Username: Robert_cheal
Post Number: 11
Registered: 08-2006
| | Posted on Wednesday, October 11, 2006 - 05:27 am: |
|
Gerald,
I accidently posted before I was finished.
I would like to make a new gantry like the one Steve posted. I've just briefly looked at your drawings at this point. So I am not sure exactly what you have in mind yet. But I would be willing to contribute to the "americanized version" when I work on my new gantry drawings. Hopefully soon.
Robert |
Gerald_D
Registered
Username: Gerald_d
Post Number: 286
Registered: 11-2005
| | Posted on Wednesday, October 11, 2006 - 07:23 am: |
|
If folk have gone over to the car thread, have liked what they saw there and feel they can have plates laser-cut to build that car, then I would suggest also to have the gantry-ends laser-cut (Like the MechMate). This was how my logic worked. Realise that the PRT y-car is probably in most need of stiffening - more so than the gantry.
An absolute key to all of this is laser-cut steel, and you being able to find someone who will cut it for you. Before you go any further, I suggest you mail this temp. drawing of the y-car to some potential suppliers to get some confidences and costs.
To complete the gantry/car stiffening exercise, have a look at these holddowns as well. |
Gerald_D
Registered
Username: Gerald_d
Post Number: 288
Registered: 11-2005
| | Posted on Wednesday, October 11, 2006 - 12:01 pm: |
|
How to attach the existing rails to rectangular tubes:
The huge bolts to the Unistrut are an overkill and too far apart. One needs to be able to shim/push/pull/tilt the rail at about every 8 to 12" when getting them perfectly set up. 5/16" bolts/screws at those intervals are perfectly adequate. The "drill" (pun intended) is as follows: (after welding the gantry)......
1. Take a 2 lengths of .75x.25 flat bar, nearly as long as the rect. tube. Mark a good screw spacing off onto one of the flat bars, making sure to miss the redundant huge holes in the rails. Center punch the one strip only.
2. Clamp the second strip to the back of the first strip (vice grips) and drill 1/4" holes right through. Unclamp, reverse/flip the strips and see if all holes still line up. If they don't, you need to mark the strips for the position in which they will finally be used - top & bottom and end for end.
3. Using the strips as templates, clamp them to the rails and tubes. Measure from side to side for parallel rails, etc. - there isn't much room for error on a 5/16" screw through a 1/2" hole. First drill 1/4" holes in all of them. Lots of holes I know, but 1/4" is easy.
4. Enlarge the holes in the rails to 1/2" and enlarge the holes in the tubes to only 3/8" (keep the tube as strong as possible.
5. Enlarge the holes in the strips for the right tapping drill size (I havn't tapped a non-metric thread in 30 years, I can't remember the size) and tap 5/16UNC threads in the strips.
6. De-burr all holes. Maybe us a big flat bar to push/scrape through the rect tube? Degrease. Paint.
For installation:
7. Push flat bar into end of gantry tube (a slot must be left there for the motor wire connectors anyway). Let the flat bar lie on the bottom. Appease your god's now...
8. At each end of the gantry, drop a short (6") length of slightly sharpened 5/16UNC "full thread rod" (Allthread/readybar) through the top of the tube and "find" the corresponding threaded hole in the flat bar. Do at both ends, leavin rods standing proud on top.
9. Lower the rail over the rods and then lift the rods to bring the flat bar up to the top surface. Add more screws. Etc. Give thanks where thanks is due..... |
Robert Cheal
Registered
Username: Robert_cheal
Post Number: 12
Registered: 08-2006
| | Posted on Wednesday, October 11, 2006 - 04:52 pm: |
|
Gerald,
Thank you for the additional information when I get comfortable with the drawings I can check with some shops for the lazer cutting. I really like your straight forward approach to finding solutions.
Gotta go now, Robert |
Gerald_D
Registered
Username: Gerald_d
Post Number: 339
Registered: 11-2005
| | Posted on Saturday, October 28, 2006 - 09:03 pm: |
|
I received a mail from Brian with this sketch:
For comparison, here is the same view (to scale) in the MechMate situation:
Maybe Brian would like to walk us through his proposal here at the forum? |
Brian Backner
Registered
Username: Allegheny
Post Number: 1
Registered: 07-2006
| | Posted on Saturday, October 28, 2006 - 09:36 pm: |
|
Gerald: Thank you for posting my sketch; I tried multiple times to upload it, but had no success.
Anyway, my thoughts are to build a gantry out of readily available structural steel shapes that minimizes flex/deflection under load.
The two weak points of the ShopBot gantry, in my mind, have been: 1) the way the x-rail carries the weight of wheels/gantry cantilevered beyond the support of the X table beam; and, 2) that the wheel carrier itself can flex as it is an unsupported leg of a steel angle.
In my drawing above, the first of these problems is addressed by shifting the point of contact to the inside aspect of the X-table beam, allowing for support of the X-rail directly under the wheels. To further support and strengthen this critical area, one could construct an epoxy bed on top of the X-beam with epoxy ala the MadVac.
The second problem is addressed by using square (or rectangular) tubing as the wheel carrier itself. This construction would be far more stable and self supporting in terms of flexing with the load placed on the carrier wheels. I show a triangular steel plate (tack welded at points ABC), but I'm not sure that it ultimately would be needed.
Perhaps the lever arms and forces created by the original configuration are unlikely to cause significant deflection and my doodling is so much nonsense. But I respectfully submit it for consideration.
Brian
Walpole, MA USA
PS:
As I was typing this, another configuration of the wheel carrier occurred to me. I hope Gerald will not get too annoyed with me if I email it to him for reposting as I have been unable to upload drawings to the forum for some reason. |
Brian Backner
Registered
Username: Allegheny
Post Number: 2
Registered: 07-2006
| | Posted on Saturday, October 28, 2006 - 09:57 pm: |
|
Finally figured out the problem with posting Jpegs!
In this version, the wheel would be supported on both sides, virtually eliminating flex. Of course, you would have to have the two interior surfaces of the channel milled flat and parallel. You would also need precision spacers (the red rectangles) machined to keep the wheels centered.
A secondary benefit is that those expensive wheels would be shielded.
Brian
Walpole, MA USA
 |
Gerald_D
Registered
Username: Gerald_d
Post Number: 347
Registered: 11-2005
| | Posted on Sunday, October 29, 2006 - 04:49 pm: |
|
Hi Brian,
In the MechMate design it was important to get strength from the gantry cross-tubes to the wheel. The wheel bolt is 30% bigger in diameter which more than doubles its cantilever strength and gives it a much bigger carrying area where it mates with the gantry. Then, there are those U-shaped plates at the end of each tube, through which the bolt goes as well. See the heavy yellow line in the drawing.......
.....that is the "flow" of the connection from the cross-tube to the back of the wheel. (Have another look at the photos at the top of this thread). Much, much stiffer than our original PRT gantry, and very unforgiving. With the old PRT it didn't matter if the x-rails were exactly parallel to each other - the gantry ends would just flex to suit. (You often see the same with the guys who put gantries up on tall 10mm thick alu plates - those plates are very forgiving of badly aligned linear bearings) When we put the first MechMate gantry on our old SB table, it was an eye-opener to see how the roller would just climb up at spots where the rails pinched together a little bit. |
20. Gantry 
SBGantryInches.zip