Who knows how to solve this piping problem with AutoCAD

[JD Mather]

Hey JD, your Private message inbox is either full, or turned off.

 

I intentionally leave it full to avoid unwanted PM from this site.

Anyone who really wants to get in contact with me finds my real email address.

[JD Mather]

I was after a general solution that would work for pipe or tubing (nominal or actual sizes) so ignoring the diameter/wall thickness elbow sizes.....

 

Here is solution with dimensions that I did use (not correct for the specified pipe).

In any case - there is only one solution for my example.

 

Pipe Run Solution.dwg

 

Tomorrow I will demonstrate technique to arrive at general solution for any pipe/tubing sizes.

[ROBP]

I was after a general solution that would work for pipe or tubing (nominal or actual sizes) so ignoring the diameter/wall thickness elbow sizes.....

 

Here is solution with dimensions that I did use (not correct for the specified pipe).

In any case - there is only one solution for my example.

 

[ATTACH]48081[/ATTACH]

 

Tomorrow I will demonstrate technique to arrive at general solution for any pipe/tubing sizes.

 

 

I won't have mine ready for tomorrow on my side therefore your's will come first.

Oh well to bad.

 

 

I did spend a little time doing it even the busy schedule of mine but had to switch to 3d after 2d see attached.

The only thing left to do in to turn the 45 deg. elbow and be able to get the exact welding position on the pipe so it all fits (and if cutting will be involved) so far so good and i think it will fit like a glove.

 

 

btw it was all done with cad on my side

Edited April 15, 2014 by ROBP

[ROBP]

As a tank designer and after 2278 drawings in my study i still question why the user is making it hard to do personally i would of opted to lower the valve connection and have it at the same height and use only a long radius elbow of the run pipe API permits low type outlets with a tombstone pad witch i point out was omitted by the op.

[nestly]

As a tank designer and after 2278 drawings in my study i still question why the user is making it hard to do personally i would of opted to lower the valve connection and have it at the same height and use only a long radius elbow of the run pipe API permits low type outlets with a tombstone pad witch i point out was omitted by the op.

 

You sound like the kind of guy that gives engineers and architects a bad name. "Convenience" isn't the only factor that determines the elevation of tanks, tank nozzles, pipe supports, and associated equipment. Furthermore, the condition described in the first post could occur anywhere, not just near a tank connection.

Edited April 16, 2014 by nestly

[JD Mather]

I won't have mine ready for tomorrow on my side therefore your's will come first.

Oh well to bad.... ...btw it was all done with cad on my side

 

I am looking forward to examining your dwg solution.

[ROBP]

I am looking forward to examining your dwg solution.

 

dwg will posted only when completed on my side if i can find the time to do it, hopefully a was able to explain better with last pictures.

 

R

Edited April 16, 2014 by ROBP

[ROBP]

You sound like the kind of guy that gives engineers and architects a bad name. "Convenience" isn't the only factor that determines the elevation of tanks, tank nozzles, pipe supports, and associated equipment. Furthermore, the condition described in the first post could occur anywhere, not just near a tank connection.

 

Architects proposes layouts and visions, Engineers does the calculations and i draw it and do shop drawings and follow the quote according to the tank codes and find solutions in order to avoid material waste and keep it within the budget.

 

Team work is always a winner and simplest is always the best option.

[ROBP]

You sound like the kind of guy that gives engineers and architects a bad name. "Convenience" isn't the only factor that determines the elevation of tanks, tank nozzles, pipe supports, and associated equipment. Furthermore, the condition described in the first post could occur anywhere, not just near a tank connection.

 

Just a reminder i was not the one whom asked for help to solve it you did.

 

Maybe i should call-in for a sick leave of absence because of my to many postings on the subject

[SEANT]

Opinion about methodologies aside, this is an interesting problem. Degrees of freedom o' plenty.

 

 

No doubt a pure mathematical approach would be a mammoth matrix effort to contend with the system of equations.

 

 

I've got an idea or two, mostly geometrical, to corral some of the variables; hopefully I can post something tomorrow to demo the concept.

[nestly]

Just a reminder i was not the one whom asked for help to solve it you did.

 

I asked if anyone knows how to solve a specific type of piping problem. You've posted 12 times now, so if you'd like to contribute something useful toward solving the problem, I'm all ears, otherwise I already told you how I feel about playing games.

[nestly]

Opinion about methodologies aside, this is an interesting problem. Degrees of freedom o' plenty.

 

 

No doubt a pure mathematical approach would be a mammoth matrix effort to contend with the system of equations.

 

 

I've got an idea or two, mostly geometrical, to corral some of the variables; hopefully I can post something tomorrow to demo the concept.

 

Whoo Hoo, the math wizard has arrived... now IT IS ON!

[SEANT]

I’m not necessarily proud of my math wizardry with this project: For the most part, I’m gearing up for a “Brute Force” programming effort. Though . . . . the parameters associated with this problem make even that a daunting prospect. The variables of rotation angles and arc length make the refinement process complex - as too large a step bypasses a solution without a clear indication of what’s needed next. Brute force works best with just one variable.

 

I’ve attached some geometry illustrating my approach. I know the end of the 45 degree elbow has to be on the green circle. I know the common tangent (red line) has to pass through the circle and land on the hyperbola. The incoming and outgoing Rays start at the ends of the tangent lines.

 

I used the program to generate the arcs in the attached. The code’s still a bit crude, I’ll attach it here when it performs adequately.

 

Do you have more Angle/Elevation examples to which I can compare results?

Demo.dwg

[SEANT]

A couple more example solutions.

Ang45Rise6Rads9.dwg

Ang60Rise10Rads15.dwg

[nestly]

I’m not necessarily proud of my math wizardry with this project: For the most part, I’m gearing up for a “Brute Force” programming effort. Though . . . . the parameters associated with this problem make even that a daunting prospect. The variables of rotation angles and arc length make the refinement process complex - as too large a step bypasses a solution without a clear indication of what’s needed next. Brute force works best with just one variable.

 

I’ve attached some geometry illustrating my approach. I know the end of the 45 degree elbow has to be on the green circle. I know the common tangent (red line) has to pass through the circle and land on the hyperbola. The incoming and outgoing Rays start at the ends of the tangent lines.

 

I used the program to generate the arcs in the attached. The code’s still a bit crude, I’ll attach it here when it performs adequately.

 

Do you have more Angle/Elevation examples to which I can compare results?

 

Sweet!, you've clearly figured out the methodology.

Unfortunately even with the first drawing for reference, I still haven't wrapped my head around how you got there. As for other examples, the variables could be almost anything where the radius is 1.5x the nominal pipe diameter, so your 2nd and 3rd drawings are perfect tests for 6" and 10" pipe respectively, so whatever you're doing, you're doing it right.

 

The only other condition I occasionally run into is when the "rise" is too small to use a standard 45, and both fittings need to be cut back (usually to the same angle)

 

I'll study your drawing some more when I'm a little more awake and try to follow your process. Thank you.

[Organic]

The only known parameters are the bend angle, and the elevation change, both of which were given in the first post. Everything else has to be calculated from those two parameters and the radii of the fittings (8" LR in this case), which was also stated in the first post.

 

It's OK if you don't know how to solve it, no one else whom I've asked knows either.

 

One more time for clarity....

 

[ATTACH=CONFIG]48017[/ATTACH]

 

I couldn't be bothered even trying to work that out. A 90 deg elbow and a 45 deg elbow would suffice. I'd move the tank if required.

 

Luckily I'm not a mechanical engineer

I did do up some custom fabrication details for a metal shop to fabricate the other day [a small component of a larger ongoing project I am working on] which will be interesting to see how it works out.

Edited April 16, 2014 by Organic

[SEANT]

Yeah, that drawing is far from self-explanatory. The crux, though, are the tangent lines (the red line, and the two rays). They describe the mating condition of arc/arc and line/arc. The arc/arc mate (red line) will always be at 45 degrees to its axis, no matter the rotation. A Cone is the 3d manifestation of that.

I’m only interested in one elevation. A cone sliced at one elevation (an elevation parallel to the axis) will produce a hyperbola. I remembered that from the “Conic Section” section of some HS math classes. I was able to analyze a progression of tangent lines as they made their way through the circle/hyperbola to find a point where a 12” radius (for example) was available.

That demo drawing leaves out much of the latter analysis – it just illustrates a way of reducing the amount of variables.

I’ll compile the routine, post that and the source code sometime tomorrow.

[RobDraw]

Well played.

[nestly]

I couldn't be bothered even trying to work that out. A 90 deg elbow and a 45 deg elbow would suffice. I'd move the tank if required.

 

There's no way to get from point A to point B using only a 45 and a 90. As you say, you'd have to change the elevation of either point A or point B. Again, neither point is necessarily determined by a tank or tank nozzle elevation, the described condition could occur anywhere in a pipe network. Also this condition certainly isn't unique to pipe fitting either, there are countless examples where there's a requirement to maintain continuity and tangency when changing angles and planes simultaneously... Solving with lines and arcs is probably as "easy" as it gets (think roller coaster) It's a general problem, not just pipe fitting problem.

[JD Mather]

... general problem, not just pipe fitting problem.

 

Bingo! Can get "close" enough for assembly through trial-and-error, but a general solution can eliminate most if not all trial and error (scrap attempts, wasted hours, parts installed under forced tension that might eventually crack and leak....or worse....).