Hi,

We have a system where we have a pump going into about 100m of mining hose, and then into 3mm wall steel pipe system. about 400mm diameter.

The mining hose is not straight, the steel section is made of straights and 90 bends.

The pump may stop and start suddenly. Should we be concerned about water hammer, or will the mining hose dampen it out.

Does anyone know the modulus of elasticity for mining hose?

Hello Wesley,

The short answer is that you will get no really useful results modeling mining hose with a beam theory piping program - not even if it is the best one.

Your problem is more akin to slug flow rather than fluid hammer. But if the hose remains relatively full the dynamic involved is the reactive force needed to change the direction of the flowing liquid at curves. If there is no structure to react the load, the hose will move - then an analysis model would have to update the (large displacement)geometry change before it could continue the analysis. This is an iterative process. Bottom line is that it is a waste of time.

It is common practice at US mines to use "safety harnesses" to restrain mining hose. There is a very good reason for this - people have been severely injured or killed by whipping hose. Be VERY careful about how the couplings are installed and use only approved, pressure rated couplings. The US Bureau of Mines has strict rules about restraining mining hose, perhaps your counterpart also has some guidelines that you can access on the web. Also, use the right hose pressure rating and regularly inspect the hose and couplings.

Best regards, John.

While CAESAR II is the best piping program ( ), I agree with John - you won't get anything useful from modeling this hose. The reasons are:

• The hose doesn't behave as a beam. Hose behavior is more akin to cable, its behavior is requires catenary equations. CAESAR II only employs beam theory (equations).
• The stiffness of the hose is so much less than that of pipe, no loads will be transmitted to the pipe.

Anytime you run into a hose, you can just stop your piping model. You basically have a free end.

Sorry guys, I must have asked my question the wrong way.

I realise that I cannot model the mining hose, and that was not my intention. I wanted to model the steel pipe section that is attached to the mining hose. the connection between the steel pipe and the mining hose is an anchor, so really whatever the mining hose does shouldn't effect the steel pipe run.

However the pressure wave from the pump will travel through the mining hose into the steel pipe run. I was going to try to analyse the waterhammer effect in the steel pipe run.

But I thought that maybe the mining hose would dampen the pressure waves before they reached the steel pipe run, in which case I wouldn't have to worry about doing a water hammer analysis?

Also, the pressure wave velocity in the mining hose is very small compared to the velocity in the steel pipe. So I assume the wave must suddenly increase in speed when it goes from the mining hose to the steel pipe? To keep the same energy I would have thought that it would have to reduce in pressure or something?

If I did assume the pressure wave made it to the steel pipe system, I need to calculate the magnitude of the pressure wave. Calculating P using the E of the mining hose gives a very small P, but by using E of the steel pipe you get a large P. Which P do I use for the water hammer analysis in the steel pipe? Do I use the P from the mining hose, because that's where the pressure wave started?

What are your thoughts?

[ September 02, 2001: Message edited by: Wesley Taylor ]

Here is some more info, in response to an email I received.

I am not concerned about the mining hose, as that is not part of my scope. This is a very small job that we have. The client (or someone else) is going to take care of the mining hose section, and we are going to design the steel pipe section.

The hose is not underground, it is not in a mine. It is on a floating plant, in a small lake. There is two barges, one with the process plant on it, and one with the mining barge on it. The process barge stays basically stationary, but the mining barge moves around sucking up slurry and pumping it to the process barge. The two barges are connected by about 100m of mining hose, which is attached to and floating on lots of little pontoons.

The flexible mining hose section meets the process barge, from there on it is a steel pipe system within the process barge.

The is an existing 600mm line, we are adding another 400mm line. The existing mining hose is restrained quite well to each pontoon, we assume the new line will be restrained in the same manner. I have been asked to analyse only the steel pipe section.

I am only worried about the water hammer pressure waves in the steel pipe system. I would have thought that the low elastic modulus of the mining hose would dampen most of the pressure waves. We have assumed the elastic modulus of the mining hose to be less than FRP pipe.

If I did apply a water hammer force to my steel pipe section, what force would I use? I would've thought that I would've used the force calculated in the mining hose, since that is where the source (the pump) is.

Hi Wesley,

OK, I think I understand now (maybe). Presumably, this is a positive displacement pump (?) and the pulsations have a significant delta P (or does it have fast closing valves? Otherwise, why the concern about potential liquid hammer?).

The overall system structural analysis (system as an irregular space frame made up of "beams") can be handled by CAESAR II and since the source of the acoustic waves is presumably the pump, the "at the pump" flow characteristics (ignoring any damping) should be available from the pump manufacturer. The design of anchors and other supports and restraints will then be handled easily by CAESAR II. The effect of pressure pulsations on wall thickness (and weld) design for components should give you some "interesting" moments.

There may be some amount of pulsation energy damping due to the circumferential elasticity of the hose but it will be difficult to quantify it. Maybe you should assume no damping because of the inexact nature of the potential acoustic waves. Have you looked into pulsation damping chambers ("pulsation bottles")? Pulsation bottles can be arranged (spaced) so that the pressure pulsations, peak-to-peak, are reacted out-of-phase by the bottles and they thereby use the pulsation energy to damp the pulsations (magic). The problem with this (bottles) is that the pump energy time-force load application will be altered and it might be necessary to do a CFD analysis to get accurate loadings at changes in direction (points of acoustic wave impact).

Is it too obvious to suggest a good anchor at the hose end of your piping (so good that the hose will break if one barge is pulled away from the other)?

Are we getting closer?

Regards, John.

[I]I work for Environmental Hydraulics Group in Canada and my boss is a designated Specialist in Ontario. We have worked previously with your colleagues at Hatch in Mississauga and Montreal. I am providing this info freely and you understand that only you are ultimately responsible for your design. Here goes:[/]

Should we be concerned about water hammer, or will the mining hose dampen it out?
There are only two ways I (and other posters) can think of it dampening the transient energy: i) by moving or whipping about; and ii) by friction within the hose, couplings or steel system components (but this only applies to the counter-surge and subsequent cycles, so it may not prevent a burst). Expansion of the pipe material should not dissipate significant energy (within its elastic range). Change of pipe material will also not change the surge energy once generated.

You would likely not be satisfied with whipping or bursting as a surge relief mechanism, so you want to make sure the system can take it long enough (many wave travels back-and-forth, testing the limits of components each time) until friction can do its work. Caution: if flow can begin suddenly into an empty pipe, the surge pressures from rapid filling can greatly exceed the Joukowski formula head you may be using.

The hose will have a lower modulus but I would need its material and/or spec to comment on that. It may be much lower than steel (of the order of 1/2 to 2/3). The majority of the piping is mining hose, the composite behaviour in terms of transmitting transient pressures may correspond approximately to "stiff" mining hose; but a computer program is required to really track and quantify this.

But I thought that maybe the mining hose would dampen the pressure waves before they reached the steel pipe run, in which case I wouldn't have to worry about doing a water hammer analysis?
If your mining hose is rubber or plastic, 100 m long may translate to about 1/4 second travel time for the transient pressure wave assuming it is anchored and fairly straight run. There will not be much dampening on its first trip and there will be no time to "react" by getting out of the way either. You should therefore analyse this.

Do I use the P from the mining hose, because that's where the pressure wave started?
You would need to use the same P for the steel pipe as for the mining hose IF the pump is the only possible cause of transients. Also, a pump will normally slow down, not stop suddenly, therefore the Joukowski equation may be an over-estimate. If a valve can close in your steel system, you need to use the larger P from the higher modulus and the Joukowski head.

To keep the same energy I would have thought that it would have to reduce in pressure or something?
Once the transient pressure wave Pt starts moving, it will essentially keep the same magnitude (less friction) as it enters the metal piping. What governs the waterhammer potential is the velocity before flow stops (steady in this case), not the Pt's magnitude. Think of Pt as a signal (Volt or traffic accident) telling a transport phenomenon to slow/stop (current or vehicles slowing as the space between them is reduced): the kinetic energy (velocity) does not immediately respond to the change in potential energy (pressure); a short time is required for this response to propagate through the system. Remember, this is transient behaviour during which the fluid is elastic!

Therefore, there will not be a sharp (local) change in velocity due to the change in material from mining hose to metal system (conductor or lane reduction): rather, continuity will ensure the effect will be felt throughout the system. E.g. the mining hose's modulus does not protect the system from Pt transmission, only reduces its magnitude when generated. A computer model is normally required to analyse changes during transmission, however if the length of the mining hose is 90% of the overall piping, the system will behave in a composite manner: like a stiff mining hose.

What are your thoughts?
If the hose can take it and flow cannot be valved-off in the metal system, you should be ok.

Recall that we consult in this and sell software, so..
I hope all this helps and I encourage all COADE users to reserve some time and budget for hydraulic transient issues before undertaking a pipe stress analysis with CAESAR II: after all, even the best pipe software in the world needs correct inputs (forces on anchors, thrusts on elbows) to produce correct outputs. GIGO is universal.

P.S. Lest anyone think me an evil spammer, it was Richard Ay who suggested I write the above reply, and it does seem that Wesley may be ok with this system – especially if he gets feedback from the operators about the surge behaviour of the existing line...

Opions expressed are my own. Facts are for you all to verify...

AMEN!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!


I wish I a dollar for each time that people mistake structural analysis for unsteady fluids analysis.

Each type of analysis will yield unique solutions and numerics. As far as so-called hammer events goes, the initial hammer wave may be dwarfed by subsequent waves if the system is poorly designed and cavitation, column separation and column rejoining occurs. The field of unsteady fluids is every bit as specialized as pipe flex analysis, and should be dealt with by competent experienced individuals only.

Thanks for all your help guys.

We have tried asking the operators about the existing line. They tell us it gets a bit of waterhammer, which they say is when they suddenly run the pump backwards (maybe to try to unclog the pump or something?). However we cannot get anyone to quantify the magnitude of the water hammer. Sometimes they said the mining hose shakes a bit, but that could also be due to slug flow type issues associated with the slurry.

There are no valves in the system.

I figured that I should use the force calculated in the mining hose, and not in the steel pipe, since the pump is in the mining hose. So thanks for confirming that suspicion. We can't find any pump starting or stopping time data, so I've made an educated conservative guess. The force is quite small, and not much of an issue.