I am interested in modelling a classic heat pipe to better know these systems. I would like to see how the temperature settles inside of it axially. Attached you will find a quick sketch: it is composed of an envelope, a "wick" which is rather axially grooved containing liquid phase, and the vapor tube.
I have modelled it the way I show in the attached file but I am stuck with the "capillary pumping" of the system. For those who are familiar with heat pipes, the capillary action is ruled by the surface tension delta(p_capillary)=sigma_liq/R_menisc or is balanced with pressure losses in liquid and vapor phases. How can I set the rotational speed of the pump? The displacement being all the volume of the wick?
Thank you very much guys. I think it is worth looking into if you are interested in these cool systems
Solved! Go to Solution.
I'm not really familiar with "capillary pumping" but may'be this could help. The displacement and the volumetric efficiency allows you to set your mass flow rate. If i understood what you need is to model your volumetric efficiency depending on the pressure difference. Am i right?
Do you have a mass flow rate target?
May be you could use an open loop to set your pump like in Step-by-step-modeling-of-a-heat-pump article and then connect it to your two-phase flow loop.
Thanks for the kind reply !
Indeed there is a pressure difference between phases that lead to the fluid pumping. The mass flow rate is given by m_dot=Q/l_vap (l_vap being the latent heat of vaporization). I tried putting sensors of it and then doing a difference as l_vap=h_vapor-h_liquid (h being the specific enthalpy) but it does not seem to work well especially at start up as both are equal during the initialization.
m_dot is around 3e-3 kg/s in steady state (calculated with Matlab) but I would like to see how the transient looks like thus Amesim
What about the heat pump ? In the compressor, it is a given value "k" am I right? I think this value should be varying in time for me, but depending on the temperature that sets up progressively in the whole system.
I've created a sketch inspired from yours in rev16:
The target of the section i added in the left side is to calculate the latent heat of vaporization at mean pressure. Then, thanks to the heat flux, i get your target mass flow rate. The rotary speed is finally deduced from the compressor/pump doc where m_tot = f(rotary speed).
Does it help you to model your system?
Thank you very much !
I have done pretty much the same thing but considered directly both h_v and h_l (for l_v=h_v-h_l) but it didn't work. But with your transformers it is ok, how come?
Also another question: my gas fraction value for the central node in vapor line is giving 0, it should not right?
I think may'be it is because the values are dynamically calculated in the transformer whereas fixed values could constrain your calculation depending on your initial conditions.
About, the oscillations, it is hard to answer to you without the Amesim model unfortunately.
Indeed, the gas mass fraction should not be 0 as you described your model but I don't have enough inputs to tell you why it is the case in your model.
For the chamber... well i created a quick model with R-C causality in all the pipes that's all
Thank you for the reply.
If you want to get pipes better I can of course send you the model; for now I can consider the topic resolved (as for the capillary pumping) ! I will give updates if I cannot solve the gas fraction issue
Me again, can someone hint me on how to do a proper "pressure and X (gas fraction, temperature, specific enthalpy... initialization" for the two-phase fluid? I have noticed that the results can be very different from one configuration to another...
I have discretized with slices in evaporator, adiabatic and condenser sections to better observe the change of fluid phase but after a while it gets wrong like the fluid is not condensed enough or the opposite: it is full vapor everywhere...
For now I am taking the same pressure for all pipes as the saturation pressure for water at the mean temperature between hot and cold source, and input the specific enthalpy (2700 kJ/kg) at this pressure in the vapor line and 0 gas fraction in the liquid line. It does not seem to work very well to be honest :/
You have to consider the charge of the fluid in your system. I suppose that your initialization of the states variables did not consider this aspect and leads to the current results. Please refer to Simcenter Amesim demo “Heat pump configuration” and especially on "closed loop - Connection of the previous model to the compressor - stage 3". It will show you the need to consider the mass of the system before closing it.
For sure if you still have an open loop, this explanation is not straightforward and maybe the key is on the heat exchanges calibration.
keep us posted ^^!
I tried setting up the initialization values more accurately by plotting and using the values (P and h) from a saturated plot for water (the bell one in log(P)-h) and it seems to work fine for the gas mass fraction now: it condenses at the condenser and vaporizes at the evaporator.
I have now another problem... my envelope capacitances show some good values but not in the right way... By this I mean that:
- I discretized my condenser, adiabatic and condenser region into several nodes (like slices)
- The adiabatic section nodes seem to have pretty good values but I expected the nodes close to the condenser to be colder and the nodes closer to evaporator warmer. The opposite appears...
I think the heat exchange between the vapor pipes (see pictures, the upper line) and the wick should be played on with the heat exchange gain should be used but I'm struggling to set the right values and feel like I am cheating by setting random numbers. Is there a way to do this properly?
Thanks! Waiting for a reply