Making round holes in rolled plate

[mstyer]

I work for a company that builds pressure vessels. The larger vessels are formed from plate that is rolled to the needed diameter. These typically have several pipes passing through the vessel wall on various chords and some radial. The challenge is to layout the proper shaped ellipse for the CNC plasma cutter to cut into the plate so that it forms a round hole once the plate is rolled. The more advanced challenge is to layout the bevels so that they may be cut before the plate is rolled.

Is anyone familiar with a formula or software to help with this task?

[brucer]

i would think about any of the current cad/cam systems could do it..

i would look at mastercam or gibbscam and solidworks.. these arent cheap, but you might be able to farm out the designing in solidworks then import it into whatever software your using to generate programs..

i'm not familiar with any of the plasma specific software though..

you would also have to have a 4 axis machine to cut it...


i'm pretty sure the old dated vesion of mastercam i have wont do it, i can look though..

[BlauSchuh]

interesting problem. i was playing around with Rhino 3D a while back and did something like this...
Here's a crude explanation from memory :)
draw cylinder surface
draw circle so the plane created by the circle is parallel to the cylinder length-wise
Use 'project' command to project the circle onto the cylinder, creating a ellipse...
'extrude' the ellipse so that it becomes a solid
get the 'boolean difference' between the cylinder and the ellipse (still looks like a circle). viola... you now have a hole
do a surface 'unroll' of the cylinder. and now you have your ellipse.

good call on the 4th axis... more like a 5 axis maybe. that bevel is doing all sorts of wierd stuff at this point.

what kind of CNC do you have?

[fortyonethirty]

Autodesk inventor has a "flat pattern" tool used for sheet metal fab. It should do the trick.

[BlauSchuh]

got bored.. but here is what your wanting... PM me if you need some assistance :)

http://i55.tinypic.com/260sprp.png

that unrolled flat surface indeed contains an ellipse...

[Fegenbush]

The key is to make the pipe pattern first, cut the pipe, then trace the outline on the vessel. If the nozzles are small (relative to the diameter of the shell) and radial, it's not a big problem to use a compass to layout the hole in the shell from the center mark. If they are tangential or at another odd angle, see the aforementioned tip of cutting the pipe first.

The flat pattern in any of the 3D parametric cad packages is great for getting the nozzle wrap. It is also a big help in laying out the repads, when they are required.

[Scott Young]

what is wrong with rolling the plate and then cutting the holes after? the plate will not roll consistently having cut out a portion. you won't have the same resistance against the slip. I am not saying it can't be done. I ran a slip roller that would roll a 3 inch thick 10 foot wide to a diameter of 13 feet. we also had a smaller one, but I don't remember the limits on it. I ran them for 2 years. where you have the cut out you will get a tighter roll.

also, you will have to be exceptionally carefull when combining the cans that your precut holes are positioned correctly. too many variable to go wrong. we would build 50+ foot vessels/tanks (that is 5 cans plus caps). I would much rather have all my cans welded before laying out for any holes. you're talking about holes that have to be inside of 1 degree of error.

[MrLeadMan]

When I worked in a tank shop we always rolled plate and then layed out and fit holes, manways and such.

[mstyer]

Thanks for all of the help so far.

To answer the question about cutting holes after the rolling process, we are looking for the most efficient way to build the vessels. The plates are already on the burn table to be cut to size. A CNC plasma cutter is much faster and more accurate than laying out a hole with a tape measure and manually cutting the hole.

Our design engineers are using Inventor and the burn table's CNC is driven by ProNest software. What I've heard is that their current ellipse shapes were laid out using the trial and error method. I'm trying to take it a step further and give them the information necessary to lay out any hole on any size pipe and take into account differences in elongation from the ID to the OD plus any bevel for welding.

[Rick V]

mstyer I can certainly understand why you want to do this; it sounds a great idea - in theory. However, Scott Young does make some important practical points based upon actual experience: "the plate will not roll consistently having cut out a portion" and "you will have to be exceptionally carefull when combining the cans".

I would have thought that information on whether this idea would work would be known to the major manufacturer's of pressure vessels. Try a search on the Internet. Folks that may have been successful in doing this would likely publish their company's accomplishment in conference proceeedings - to show off their expertise and attract potential customers.

This is the type of search that would be turned over to a company librarian with the skills to access all the major databases and search all applicable conference proceeedings.

If you find nothing, your next step would be to proceed to a simple trial - using left over material - plan a special test piece, make your ovals, etc and then roll it and see what you get.

[zipzit]

I may be a mediocre welder but I'm pretty good at writing software. Check out the tubenotcher program at

http://www.cobratorch.net/ttn_dxf/

When doing work on big stuff, obviously you can't use your 8.5 x 11 printer. So I added a .dxf output option, a format compatible with most CAD systems. Input the diameters of the pipes / vessels involved, in this case enter the large diameter vessel as "cut tube" the small diameter pipe is the "base tube". Program also handles angle, material thickness, and offsets. DXF out, then capture the data you need to import to your CNC program. Save your file as "Filename.dxf" (Never could fix that suffix thing.) I was careful that all the key lines of output are on seperate layers in DXF format. Oh, yeah, and its free. .

If you need a software tweek or something customized, its a discussion. PM me.
--zip

[Knotbored]

To avoid the rolling distortion perhaps the "plug" could be left in the hole with 4 tabs that would be cut out after rolling.

[Scott Young]

all I can say is try it. you will find the the plate is distorted in the rolling process. your hole will go though the distortion and may move on you. I don't know for sure as it is not the most efficient practice. I do know I have rolled repads (reinforcement pads) that have been precut and at times the hole needs to be trimmed alittle due to distortion. I don't understand how you think it is more accurate to cut the holes prior to rolling and fitting. I suggest calling or visiting several plants that make vessels and tanks and see for yourself. I have worked for two tank companies and have been subbed out to another as a roll operator/fitter/welder. What you are wanting to do isn't done. Now I am not saying it can't be done or it isn't done elsewhere. All i can say is cut some holes, roll a plate, and make a tank. That is a big gamble considering the cost of that plate.

We have the fitter that lays out the hole. We have the project manager confirm the layout with the prints. We have the QC sign off on the hole placement. Then, the hole is cut. After the hole is cut, the the nozzle or manway is placed and fitted. Manger and QC again sign off on placement/fitting and then the welder begins welding. I don't believe you can improve upon the accuracy and efficiency.

As for the trial and error part of laying out the ellipsis, you need to fire your fitter and hire a fitter than can fit tanks. A good fitter can lay out a tank to the 1/32 of an inch and better. A good fitter can draw and make his own templates that he can use to transfer to the tank for difficult placement. most of the time a good fitter can layout the tank with his appropriate tools designed for laying out ellipsis and circles.

[Fegenbush]

Here's my anecdotal story from an old coworker:

They hire in a "layout man" at a time when the shop is full of work. This guy comes in and spends a few days getting acquainted with the shop practices and types of vessels and kettles they make. He then walks into the shop office and asks the super (who is the guy telling me the story):

"What's a layout man 'spose' do?"

The super says to the guy "A layout man punches his time card and gets the hell out of here."

Back to reality. What you ask, the layout of nozzles and manways on a rolled shell is *exactly* what a layout man 'spose' do. There are many options for using your CNC equipment to assist in the process, and I would be happy to discuss them with you via PM, but there is no clear cut, easy answer. If there were, there would be no need for layout men. A good layout man is worth his weight in Hastelloy X.

[CADJockey]

Hi all -

Scott Young is correct. Sometimes automating is not the best way to go, and we find that shortcuts can cost us more time/money.

I worked on the fab floor of a large structural steel house as a welder, then a fitter for ten years. My next gig I served a 4 year apprenticeship in the SMWIA and became a journeyman sheetmetal mechanic in a heavy guage shop, which I did for quite a while. In both of these engagements I did a lot of work on a variety of large round vessels; both pressure vessels and a lot of millwright work for grain elevators, etc. I also became skilled in making rotational molds, both large and small. I can tell you also that I have spent the last 20 years as a design draftsman with a variety of AutoDesk products at my disposal, including AutoCAD and Inventor.

Here's what I think I know - 9 out of 10 guys who say they are fitters couldn't lay out hopscotch on a sidewalk. Doing this type of fitup work requires a lot of skill, both in math and in common sense and previous experience. I suspect this is why mstyer is looking for a way to burn these holes on the burn table. And it would be faster, IMO, if it would work well. But for the reasons Scott Young has already given, it is not a practical solution. Plate or sheetmetal will distort in the rolls at the places the cutouts are. I've done it that way. I have also tried the trick with leaving the blanks in place with unburnt tabs and, while it helps, it can be problematic. The big thing is definitely having your cutouts end up where you intended them to be according to design and/or print. Another big issue is the bevels involved on the holes burnt in the plate/sheetmetal. Maybe things have changed since I worked on the fab floor but in my time our burn tables and plasma tables were only capable of blowing a kerf that was perpendicular to the material. In a development of the required hole, unless the hole is perpendicular to the c'line of the vessel, there is going to be some angle of the edge as it passes through the tank wall. If it is very thin material, it is not so big of an issue, but the thicker the material, the more pronounced the issue becomes.

I learned my pattern layout and development and I became a good layout man. There is no substitute for a guy who can develop a pattern, whether it be on the computer in CAD, or on a piece of brown butcher paper out in the shop on the layout table. And also, some skill of hand is also required with a cutting torch/plasma torch and your grinding tools. I can do these holes in Inventor and unfold a section into a flat pattern to burn on a table, but I still believe the best practice is to make your holes AFTER the tank walls are done.

[fdcmiami]

i worked as a layout person for a large industrial ventilation company years ago (blowpipe)

i also did most of the plate work. it was always easier to fit the pieces and cut the holes after rolling. most skilled pattern makers, at least those familiar with triangulation and parallel line development could layout the pattern of the hole from the pattern for the fitting. it was not practical in most cases to do so for the reasons desribed above, light gauge stuff, yes, plate no.

here are some patterns and some mathematics from back in the days i was a pattern maker. i did not include cone on cone as it was not discussed.

lol paper is starting to take on the appearance of parchment.

[Jimmer]

My first job as a designer for a pressure vessel company was to figure out how to do that for their big filter vessels. It's a snap in Solidworks, model the rolled body, extruded cut the holes, flatten the sheet, and send that flat patten out to plasma. It saved our fitters hours of time cutting the holes.

The trick was to have the plasma guys "stitch" cut it, leaving it attached, then we had the rolling guys roll the shell, weld the long seam, then our fitters would just cut the "stitches" with a hand held plasma and insert the pipe.

[Awelderiam]

I dont really see all of the issue with the elongated holes......on all of the large diameter vessels the shop will back gouge and weld from the inside anyhow so what is the big deal about perfect fitup anyhow? A 1/4" gap on a 3" thick vessel is nothing.......JMO. Now on the smaller diameter when it is a full penatration weld from one side then thats where this comes into play....but 90% of those are piping and the CNC plasma's can cut every hole with bevels all on Cad drawings..........

[firestick]

The key is to make the pipe pattern first, cut the pipe, then trace the outline on the vessel. If the nozzles are small (relative to the diameter of the shell) and radial, it's not a big problem to use a compass to layout the hole in the shell from the center mark. If they are tangential or at another odd angle, see the aforementioned tip of cutting the pipe first.

The flat pattern in any of the 3D parametric cad packages is great for getting the nozzle wrap. It is also a big help in laying out the repads, when they are required.

I love you guys where were you 20 year ago when I needed you .
Mr Fegenbush 537976 sounds like he has some years of experiance in the job.
You can't beat the dividers and a centre pop trial run ,the tip of cutting the pipe 1st and tracing outline on the vessel shell for starter is a quick way out if your stuck in the boonies. You didn't mention the material did you, if it's stainless get some more advice before you ripp a hole in the bank balance.
FIRESTICK THE AUSSIE WELDER .