Gradual bends in underground PVC

Status
Not open for further replies.
I once had to lay 1" PVC in a ditch that looked like a drunk, blind man dug it. It was so wavy it would make you nauseous just looking at it. Anyway, when all was said and done I think I had at least 8 or 9 90-degree bends in it. Pulling 3 #10's in it was extremely difficult with 2 men pulling and me feeding.
 
peter d said:
I once had to lay 1" PVC in a ditch that looked like a drunk, blind man dug it. It was so wavy it would make you nauseous just looking at it. Anyway, when all was said and done I think I had at least 8 or 9 90-degree bends in it. Pulling 3 #10's in it was extremely difficult with 2 men pulling and me feeding.
Click to expand...

I guess the important take-away is that large bend radii do not help reduce pulling tension. They reduce sidewall pressure, but the pulling tension multiplier of each bend is a function of angle and friction coefficient, and not radius.
 
Carultch said:
I'm looking at the famous "360 degree rule", and I'm thinking about uneaven trench topography and very gradual bends in the run. I'm talking so gradual, that you can make the bend with the room temperature flexibility of the conduit. Possibly 40 ft radius and greater.

Is there a radius at which these bends would no longer count toward the "360 degree rule"?
Click to expand...

Most replies here are from experiences that are on relatively flat terrain.

Some may feel differently if they ever do have a really long run - say 1500 feet or more between access points and in pretty hilly terrain.

I have done those before for irrigation applications, though we usually lay the wire out and slide the conduit over it and then bury it, or use "cable in conduit" products. Would hate to try to pull some of those. If you measure length of run using point to point GPS methods you don't account for all the changes in elevation along the route and will be short on materials.
 
Carultch said:
I'm looking at the famous "360 degree rule", and I'm thinking about uneaven trench topography and very gradual bends in the run. I'm talking so gradual, that you can make the bend with the room temperature flexibility of the conduit. Possibly 40 ft radius and greater.

Is there a radius at which these bends would no longer count toward the "360 degree rule"?
Click to expand...

around here, for SCE anyway, 12' radius doesn't count as a 90.

less than that does, i do believe. on the other hand, SCE only
wants you using 12' radius in the ground, except to come up into
gear. so no runs ever have more than two 90's in them.....



theoretically. :angel:
 
Ultimately, as Carultch pointed out, pull tension and sidewall pressure are what matter. The Code requirement is just a way of keeping those within reason without having to do a more complex calculation.
 
Carultch said:
I guess the important take-away is that large bend radii do not help reduce pulling tension. They reduce sidewall pressure, but the pulling tension multiplier of each bend is a function of angle and friction coefficient, and not radius.
Click to expand...
The guys with the calculators always say this, but my hands and my shoulders tell me different. It's harder to pull through a three point saddle or back to back nineties than it is a long, gentle bend even it shaped like a horseshoe.
 
Smart $ said:
Ultimately, as Carultch pointed out, pull tension and sidewall pressure are what matter. The Code requirement is just a way of keeping those within reason without having to do a more complex calculation.
Click to expand...
I agree with that.

ActionDave said:
The guys with the calculators always say this, but my hands and my shoulders tell me different. It's harder to pull through a three point saddle or back to back nineties than it is a long, gentle bend even it shaped like a horseshoe.
Click to expand...
Raceway type, conductor type, and even temperature can make a difference.

I've particularly noticed that any XLPE conductor types being pulled into PVC seem to pull much easier when it is cold outside then when it is hot outside. I have also notices steel raceways pull easier then same conductors being pulled through same sized PVC.
 
kwired said:
Most replies here are from experiences that are on relatively flat terrain.

Some may feel differently if they ever do have a really long run - say 1500 feet or more between access points and in pretty hilly terrain.

I have done those before for irrigation applications, though we usually lay the wire out and slide the conduit over it and then bury it, or use "cable in conduit" products. Would hate to try to pull some of those. If you measure length of run using point to point GPS methods you don't account for all the changes in elevation along the route and will be short on materials.
Click to expand...
We used to lay out the conduit glue and suck a stringline thru then pull back rope and consecutively larger pulling cable. Hook cable to Nissan, pull wires, then stick large sweeps on and finish each end. Wiggles of conduit were never a concern. Intententional 30, 45 were.

Now, the guys doing cablecon have it in and buried before we could lay out the conduit.
 
ptonsparky said:
We used to lay out the conduit glue and suck a stringline thru then pull back rope and consecutively larger pulling cable. Hook cable to Nissan, pull wires, then stick large sweeps on and finish each end. Wiggles of conduit were never a concern. Intententional 30, 45 were.

Now, the guys doing cablecon have it in and buried before we could lay out the conduit.
Click to expand...

I hear you, especially those that can plow it in, they don't even need time for backfilling.
 
iwire said:
You can choose not to count just as easily as the inspector can choose to fail it if it exceeds 360. :cool:
Click to expand...

Yes, well, I'm not an AHJ, but I was just saying if I were...

Probably my practical rule of thumb is that if I have to put any bending pressure on the conduit to get it to fit into the trench, and I can't afford to count that in my 360, then I'm going to ask whoever dug the trench to straighten it.
 
kwired said:
Raceway type, conductor type, and even temperature can make a difference.

I've particularly noticed that any XLPE conductor types being pulled into PVC seem to pull much easier when it is cold outside then when it is hot outside. I have also notices steel raceways pull easier then same conductors being pulled through same sized PVC.
Click to expand...
Pulling any kind of conductor through any kind of pipe is harder to do through back to back nineties than a series of gentle bends any day of the week any season of the year.
 
ActionDave said:
The guys with the calculators always say this, but my hands and my shoulders tell me different. It's harder to pull through a three point saddle or back to back nineties than it is a long, gentle bend even it shaped like a horseshoe.
Click to expand...

That is the mathematical model. The friction from going through a turn depends only on the angle of the turn and the tension at either end of the cable. Furthermore it is exponential with the angle of the turn, and multiplies the tension on the cable going in to the turn. Anything you can do to reduce tension going in to the first turn will make a huge reduction in tension at the end of the pull.

As you note, reality deviates from theory. I think that the best use for theory is to give you an up front approximation of pulling forces, and also to help figure out which end of a run to pull from.

Jon
 
winnie said:
That is the mathematical model. The friction from going through a turn depends only on the angle of the turn and the tension at either end of the cable. Furthermore it is exponential with the angle of the turn, and multiplies the tension on the cable going in to the turn. Anything you can do to reduce tension going in to the first turn will make a huge reduction in tension at the end of the pull.

As you note, reality deviates from theory. I think that the best use for theory is to give you an up front approximation of pulling forces, and also to help figure out which end of a run to pull from.

Jon
Click to expand...
Thank you.
 
wwhitney said:
Exponential, wow. What's the physics behind that?

Cheers, Wayne
Click to expand...

Dunno exactly, but a 1* turn has virtually zero drag whereas a 180* turn is near infinite-the point load at the apex of the corner doesnt rise in a linear fashion.

Good wire lube and lots of it are the second best way to make pulls easier/possible;first is use as few bends as possible.
 
wwhitney said:
Exponential, wow. What's the physics behind that?

Cheers, Wayne
Click to expand...

A long time ago I was walked through the derivation, but I can't pull it out of my hat.

The first friction approximation that sliding friction is simply <coefficient of friction> * <normal force>, and doesn't depend upon the area through which the force acts.

Next you ask 'what is the side force when a cable changes direction?' Ignore a continuous curve, just look at a turn at a single point. It is pretty clear for a turn around a single point is just the vector sum of the two cable tensions on either side of that bend. This gives the basic point: at a bend the friction force depends on the tension of the cable on either side of the bend. Note the approximation that the weight of the cable is being ignored.

Next step: to pull a cable the pull force at any point must exceed the friction at that point. Looking at the bend, the two cable tensions are: the tension on the entry side of the bend (call it A) and the tension on the exit side (the side you are pulling toward, call it B) of the bend. B = A + <bend friction> and <bend friction> is a function of A and B. The greater A is, the greater the side force and thus the greater the friction adder.

To finish the derivation you'd need to treat a continuous curve as a series of point bends, and then take the limit as the number of bends is made infinite while the angle of each is made flat. I'll leave that as an exercise for the student (meaning I don't remember how to do it ;) and point you to the answer:

http://www.polywater.com/techtalk1.pdf

For a bend: Tout = Tin * e^(<mu><theta>) where <mu> is the coefficient of friction and <theta> is the angle of the bend in radians.

The exponential nature of the friction going around a bend is put to good use controlling things like the speed of ropes under load, or the winches on boats; the rope is wrapped around the winch, and the friction between the rope and winch is adjusted by small forces one length of rope, with much larger forces being controlled on the other side of the winch.

-Jon
 
It does not matter except for sharp bends, but at some point you need to take into account the force to bend the conductor itself as it goes into the bend and out again. For solid wire it might actually be significant.
 
Gradual bend, has a "gradual band" been defined? As would be that a transformer is hot without providing an actusl temperature. These are relative terms which are left to your opinion based upon experience.
The concern to have would is if the conduit dimensional was compressed where the conduit was no longer round and squeezed which increases as the bend tightens.
If I were to guess there would be abrupt bends that would squeeze the conduits dimensions so my opinion is that the the bends a refered to my the OP are of no consequence and much to-do about nothing
Common sense we would be best to use.
 
templdl said:
Gradual bend, has a "gradual band" been defined? As would be that a transformer is hot without providing an actusl temperature. These are relative terms which are left to your opinion based upon experience.
The concern to have would is if the conduit dimensional was compressed where the conduit was no longer round and squeezed which increases as the bend tightens.
If I were to guess there would be abrupt bends that would squeeze the conduits dimensions so my opinion is that the the bends a refered to my the OP are of no consequence and much to-do about nothing
Common sense we would be best to use.
Click to expand...

I defined "gradual bend" when I opened this discussion, as a bend that is made with the room temperature flexibility of the conduit. I.e., no factory elbows or sweeps are used, no 5 degree couplings, no bending equipment that applies heat to make it bendable, and no bending equipment of any kind.

Pick up a piece of 1" PVC conduit, and observe how it bends under its own weight. Now with a partner, see how much curvature the two of you can bend the conduit by human forces alone, and without intentionally heating it to make it bend more. That's what a "gradual bend" is.

If desired, we could consider a radius of curvature for this, such as 40 ft, to make a more objective definition.
 
Status
Not open for further replies.
Top