Calculating Friction

[Eleventy]

I am a recent graduate who recently got hired by a commercial irrigation company. Got hired, did a bunch of manual labor etc.. to learn the industry and now my boss has me learning how to do the calcs etc... for how to design these systems.

I suppose I should qualify this really quickly, neither my boss nor his boss (the owner) are engineers. I mean, we are in kansas, so anybody could potentially be an "engineer", but the owner is an accountant and my boss is just a very sharp guy who never went to college and got promoted to his position by his abilities. They both realized, however, that having somebody with a solid understanding in math would be beneficial as their company expanded and hired me at a bargain basement price (haha, which I am a little bitter about but I had to take an offer asap to pay back my loans).

Anyway, going through my boss's spread sheets I realized he was using the Hazen-Williams Equation to calculate friction loss. I know that the Darcy-Weisbach equation is the more "correct" equation but that it might be more difficult to impliment into a spreadsheet due trying to pull values off of the Moody Diagram. However, all the irrigiation water around here is well water and the pipes are all buried so the temperature of the water would remain constant as it goes through the pipe correct?

My boss has been away for about a week and he basically told me to go through and learn how to use his spread sheets etc.. and come up with a list of questions for when he gets back.

SOP according to the irrigation books I have read is to use the Hazen-Williams equations, but what is SOP in other industries? Does it matter? I am probably going to try to make a spread sheet for comparison but I was hoping some people with much more experience and knowledge in this area could help alleviate my ignorance.

 

[katmar]

Have a look in the FAQs tab at the top of this page and read quarks FAQs on friction factors for the Darcy-Weisbach equation. This avoids reading the Moody diagram. You will find many engineers here have settled on the Churchill equation. Run it and see if it is significantly different from your H-W results.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[BigInch]

Go with churchill until your actual pipe pressure and flow readings force you to modify it.

"We have a leadership style that is too directive and doesn't listen sufficiently well. The top of the organisation doesn't listen sufficiently to what the bottom is saying." Tony Hayward CEO BP
"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.liv

 

[pipesnpumps]

I am going to recommend you stick with H-W. You are on a process industry forum, and these guys live for Darcy Weisbach, probably even use the two-K loss coefficient method..

You don't need any of that. If your boss already has the C factors for the PVC fittings and pipe, and equivalent lengths, then stick with that because the difference in the formulas won't affect the results.

Learn to link the spreadsheets together for different diameter runs, maybe spiff up the spreadsheet with drop down selection boxes (for example drop down box to pick diameter, or types of fittings). I would work on your people and business skills while in this job, then take the first opportunity that comes along.. It is tough out there right now, good luck

There are probably a blue million darcy weisbach spreadsheets out there available for downloading. I recommended buying a copy of Crane 410 manual if you want to create your own.

 

[darai]

Eleventy (Mechanical):

You need some tough love! You should be happy you have a job, and a good job too! So get over this crap about your bosses qualifications and get on with the task you were given

Learn to use the spreadsheet! When he gets back and gives you an assignment that requires the use of the spreadsheet, what will be your excuse for not knowing how to use it?

Do not become a know-it-all all of a sudden, else you'd be on the dole in a New York minute!

So just pass the first test before you; learn to use the spreadsheet!

 

[Compositepro]

Darai, that was a little harsh. It seems to me that eleventy is do just what his boss asked him to do, and in the process has a legitimate question which is better answered here than by his boss.

 

[BigInch]

Well then let him try to keep it for awhile. What's wrong with improving something that probably needs it? How big are those irrigation pipes, what's the typical velocity, and is the water cool?

http://www.eng-tips.com/viewthread.cfm?qid=165987&page=37

"We have a leadership style that is too directive and doesn't listen sufficiently well. The top of the organisation doesn't listen sufficiently to what the bottom is saying." Tony Hayward CEO BP
"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.liv

 

[SNORGY]

Eleventy:

How accurate do you truly need to be? I have found that I can get quite accurate results by setting up a Newton-Rhapson iteration routine for the Colebrook equation to determine the friction factor. You may have to return absolute values of the difference between Fn and F(n+1) and pick a low positive value for your initial seed number; I discovered years ago that the shape of the curve is such that you should always be able to converge to the correct value if you do this.

Failing that, I have found that for Reynolds numbers less than 100,000 (depending on how much less), you can "eyeball in" a f=0.025 (or some such number) pretty easily, and for higher numbers, f=0.015 (or some such number). You could set up a look-up table following this approach.

I have never used Hazen-Williams myself, but I have never heard of anything completely going haywire if you do use it.

Regards,

SNORGY.

 

[Eleventy]

Firstly to darai: I wasn't attempting to besmirch my boss or complaining about my job, haha although I do like to complain about my pay Seriously though, how does me looking through my boss's spreadsheets necessitate that I do not know how to use spreadhseets? I understand your desire to give "tough love" to somebody who is beginning their professional life, however it is likely better doled out in more accurate amounts. I do agree with not being a "know it all" though.

Sizes of the pipes vary. Volumetric flow rates depend on the the well/size of the field/desire for future expansion of the system. Mainline is ~8" to 12", sub mains are ~4" to 8", the "drip tape" that runs under the field varies greatly depending on soil content/length of field etc.. and we use manufacturer's programs to calculate pressure/flowrates.

Basically here is how a system works: water comes out of well gets filtered etc... then gets piped underground through various fittings to tape inlets where it flows through the tape to the flush mains. Once the system pressurizes water starts to flow out of the tape onto the field where it waters crops from the roots up. So we know the pressure at submain manifold and at the flushmain at the other side of the field but since everything else is underground you can't measure the pressure anywhere else.

However, how accurate the readings need to be depends greatly on our customer's needs and some farmers are fine with just getting water somewhat evenly put on their entire field whereas GE seed companies want to know exactly how much is getting put where (with some % of error). The system is kind of "reversed" engineered since we want our inlet pressure at the drip tape to average some X psi and a variation in this pressure is going to effect exactly how much water is being put on some portion of the field that the tape waters.

 

[BigInch]

At this moment, I seriously doubt you can find a better overall or more accurate equation than Churchill. No iterations are required, and it works for laminar or turbulent flow, no guessing of C factors (roughness is relatively easy to find), any viscosity, so its extremely versitile and very easy to program. Anything else is 70 years of no progress.

"We have a leadership style that is too directive and doesn't listen sufficiently well. The top of the organisation doesn't listen sufficiently to what the bottom is saying." Tony Hayward CEO BP
"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.liv

 

[Latexman]

Making a spreadsheet for comparison is an excellent idea. A side-by-side comparison is especially powerful. I'd recommend getting some experience looking at these comparisons (several jobs?) and getting knowledgeable on why any differences.

Good luck,
Latexman

 

[pipesnpumps]

Eleventy, don't worry with learning the above info posted about Darcy-Weisbach and friction factors yet..

Hazen-Williams is definitely the way to go with water. Hazen Williams is very accurate all the way up to 25 ft/sec velocity in the piping, and you are definitely going to be below that for sprinkler piping.

Of course if you want it to be real accurate your spreadsheet needs to account for when the sprinkler is at the 'dead end' of a flowing pipe vs just tee'd off a main.. The velocity adds to the pressure available to the sprinkler that is at the dead-end.. Your bosses spreadsheet may or may not have accounted for this.. I bet not one person here every adjusts their calcs for it, even if they are obsessed with calculating friction factor out to several decimal places. The reason being is that it makes the calculations cumbersome.. It is just something small to keep in mind if you are getting high line velocities and need very accurate calcs.

You might consider an option on the spreadsheet that asks "sprinkler on end of line?" and if it is, the spreadsheet accounts for the additional velocity pressure available for pushing water out of the sprinkler (Pv = 0.001123 * Q^2 / D^4, Pv=velocity pressure psi, Q=flow gpm, D=inner diameter inches)


This is very similiar to hydraulic calcs for a fire sprinkler system, and you might start this discussion over in the NFPA forum.. BTW you aren't calculating it "backwards". The calcs should always start at the known pressure.. So in your case its 0 psig (ambient air) + X psig required across the nozzle + line losses + elevation difference.

As I said probably the most important thing is that your boss has already gathered the pipe roughness (in HW term that is called C-factor), and the losses for the fittings (in HW its called 'equivalent length', in D-W it is either 'equivalent length' or 'K-factor').

 

[katmar]

@pipesnpumps - you would have lost your bet. There have been MANY discussions here about the conversion of velocity head to pressure head, particularly with reference to sparger and manifold design. And you are wrong about this conversion occurring particularly at dead-ends. It occurs every time there is a velocity change and this occurs at every nozzle (not only dead-ended nozzles) and also at pipe diameter changes. If you do a search for "Bernoulli" in this forum you will see that the members here have a lot more theoretical AND practical knowledge than you give them credit for.

@Eleventy - You have stated that you are a young engineer starting out and that you are keen to learn. I admire you for that. Take my and Latexman's advice and compare H-W against D-W. This is the way you learn and develop your experience and judgement. Add in the Manning formula as well while you are at it. There is no doubt that D-W is the most accurate and flexible of the methods, but H-W is still mandated by some authorities for historical reasons, so learning it is not a wasted effort.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[Eleventy]

@pipesnpumps - our company doesn't do too much sprinkler/center-pivot design and its all sub-surface drip for corn/soybean; however I appreciate the advice.

to everybody else: thank you, I mentioned this all to my boss and he encouraged me (when I have downtime) to create a spreadsheet comparing the different calculated pressure losses for different set of equations and then I am planning on creating a database so we can compare what the calculated pressures would be for separate projects with the as-built pressure losses with some % error for construction differences.

 

[BigInch]

If you do subsurface and drips, I would imagine that you have lots of nonlaminar and transition region flows. Can't beat Churchill when you are likely to have flows in all regions.

"We have a leadership style that is too directive and doesn't listen sufficiently well. The top of the organisation doesn't listen sufficiently to what the bottom is saying." Tony Hayward CEO BP
"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.liv

 

[DLite30]

It may be worth a little effort to have an alternate method to compare designs, especially if you can show that you can reduce the size of the pipe required to flow the same amount of water at an acceptable pressure loss, saving costs.

However, I would be a little hesitant about this, as some people may take offense to you showing them that what they've been doing for years is "wrong".

Maybe the first thing I would do would be to try to clean up the spreadsheet and maybe do something in VB with check boxes, etc.

 

[rconner]

While some engineers obviously have their preference for specific design tools, I guess influenced by what they have been taught or what they have experienced, I don’t think there is anything necessarily wrong with looking at the results of different design approaches side by side. I believe that is precisely what ASCE Fellow A.M. Friend did in a hypothetical (arguably some typical?) pipeline example now near half a century ago, in the reference I mentioned in my 13 July ’06 11;19 post at

http://www.eng-tips.com/viewthread.cfm?qid=159749

. In his/that particular conditions/comparison, I believe Friend found little practical difference in the results.
While that is not to mean that there wouldn’t be any significant variations with different conditions, it does perhaps tend to tone down some partisan rhetoric (see also past discussion at

http://eng-tips.com/viewthread.cfm?qid=165987

etc).
[And then again, there might also be value to going to the field occasionally and with instruments measuring/approximating however possible what is actually being achieved – this could also vary with different conditions. Let us know what you find.]

 

[TenPenny]

When you step back and think about it, you would hope there would be little or no difference between the methods. If there is any significant difference between the methods, that would imply that one is correct and the other isn't.

 

[BigInch]

Real problems occur in laminar and transition ranges.
Other variations are also encoutered, presumedly because of roughness values changing with age of pipe. Roughness doesn't always increase with age of pipe and some pipe actally becomes smoother with age in some circumstances. The viscosity of water at 32F is 7 times its value at 212F. Why program an equation that immediately puts you at a disadvantage by not having the capacity to handle commonly variable temperature conditions, and why even write a program for an equation if it only works in rel few specific cases, or is more complex to program, yet gives no increase in expected accuracy, or even gives actual errors in others.

You don't have to irrigate in winter would seem like a lame excuse.

"We have a leadership style that is too directive and doesn't listen sufficiently well. The top of the organisation doesn't listen sufficiently to what the bottom is saying." Tony Hayward CEO BP
"Being GREEN isn't easy." Kermit

http://www.youtube.com/watch?v=hpiIWMWWVco

http://virtualpipeline.spaces.liv