Reports 1-1 of 1 Clear search Modify search
AOS (Baffles & Beam dumps)
keiko.kokeyama - 15:23 Tuesday 14 July 2020 (14742) Print this report
Estimation of thermal conductivity of the HPBD (heating case)

----------------------------------- Summary -----------------------------------
Similar to klog 14740, here is the thermal conductivity estimation using the heating curve of the high power beam dump. Again, the heat occurred at the dump is conducted from the dump to the copper flange through 10 copper heat sinks. The fitted thermal conductivity is 1400 W/m/K, where as the thermal conductivity of copper (heat sink) is 400 W/m/K around 30-60 deg C. This is consistent with the cooling measurement (klog 14740).

----------------------------- The fitting function T(t) ----------------------

Heating is done by 10W laser continuously hitting the beam dump. In this case, the differential equation of the heat can be written as
 
      M*C0 * dT(t)/dt = lambda * A / dx * (Tb - T(t)) + Pin
 
where lambda is a thermal conductivity of the path, A is the area of the path, dx is the length of that path, and Tb is an albtrary parameter of temperature, respectively. Again solving this equation, we get
 
      Tt = Tb + Pin/K + (T0 - Tb - Pin/K) * exp (-K/C *t);
where K = lambda * A/x, C = M * C0 with the dump weight M and specific heat of heat sink C0. The equilibrium temperature after a long time is thus Tb + Pin/K.
Heating is done by 10W laser continuously hitting the beam dump. In this case, the differential equation of the heat can be written as
 
      M*C0 * dT(t)/dt = lambda * A / dx * (Tb - T(t)) + Pin
 
where lambda is a thermal conductivity of the path, A is the area of the path, dx is the length of that path, and Tb is an albtrary parameter of temperature, respectively. Again solving this equation, we get
 
      Tt = Tb + Pin/K + (T0 - Tb - Pin/K) * exp (-K/C *t);
 
where K = lambda * A/x, C = M * C0 with the dump weight M and specific heat of heat sink C0. The equilibrium temperature after a long time is thus Tb + Pin/K.
Heating is done with a 10W laser continuously hitting the beam dump. In this case, the differential equation of the heat can be written as
 
      M*C0 * dT(t)/dt = lambda * A / dx * (Tb - T(t)) + Pin
 
where lambda is a thermal conductivity of the path, A is the area of the path, dx is the length of that path, C0 is a specific heat of the heat sink (ignoring the temperature dependency), M is a weight of the dump, and Tb is an arbitrary parameter of temperature, respectively. Again solving this equation, we get
 
      Tt = Tb + Pin/K + (T0 - Tb - Pin/K) * exp (-K/C *t);
 
where K = lambda * A/dx, C = M * C0 for simplicity. The equilibrium temperature after a long time is thus Tb + Pin/K.
Images attached to this report
Comments to this report:
shinji.miyoki - 22:02 Tuesday 14 July 2020 (14748) Print this report

400W and 1400W are so different with each other.

Is this system consider the finite cooling effect of copper plate heat anchor on the vacuu flange that is connected with the HPDB in the vacuum tank ?

In other word, the temperature of this copper plate heat anchor will also increase from the initial temperature. Of course, this heat anchor has lower temp due to contact with air, however, not so good cooling effect.

Two heat transfer equations are necessary 1)  for HPBD and 2) for a copper plate on the vacuum flange that has heatup from HPBD and cooling due to air circulation.  

keiko.kokeyama - 12:11 Wednesday 15 July 2020 (14757) Print this report

Do you mean 400 and 1400 W/m/K? If so, as written in the post, this difference suggests the additional heat path which is the SUS pedestal to the table. And actually this path flows the heat more than the heat links. It is the concequence of the heating / cooling test.

The model assumes only the heat flow from the dump through the heat links to the flange and the dump and flange become the commom equilibrium temperature after a long time. There is no assumption of (positive) cooling effect at the flange but I don't think it is necessary. Heat is kept sucking over the temperature change from the dump that is the cooling. It is shown on the right term of the equation of this post, lambda*A/dx * (Tb-T(t)).

Search Help
×

Warning

×