In order to evaluate the expected performance of the thermal control subsystem on the spacecraft, I performed orbital simulations of the spacecraft in worst case hot and cold conditions. For these simulations I used the NX Space Systems Thermal software from Siemens. The following sections outline how I developed my thermal math model and simulation inputs for both cases, as well as the results of these simulations.
These simulations have been performed in phase C. This page of documentation has not been changed since the Pre-CDR. However, I did receive some feedback about the simulation set up in the CSA comments that will be incorporated in the Phase D simulation.
In these sections I describe how I developed the thermal math model of the Iris spacecraft for use in orbital simulations. I focus on the meshing and material properties assigned, as well as thermal couplings applied between surfaces. All details of the model in this section apply to both the worst case hot and worst case cold simulations.
I created the meshes for the model based off of the CAD assembly of the entire spacecraft. I deleted a number of smaller parts from the assembly, such as fasteners and connectors, as these parts have a relatively low thermal mass but would significantly increase the number of elements and nodes in the model. I also simplified the CAD models of the parts by removing small features like holes to simplify meshing. Fasteners were replaced with conductive couplings between faces, which are discussed further in a later section. The components that are meshed and included in the simplified model are detailed in the following two tables, separated based on whether they are 2D or 3D meshes. In these tables, I specify the material and surface finish that was assigned to each mesh. If a component is given an 'N/A' surface finish, it is because the surfaces are not included in any of the radiation requests.
The following table gives the important thermal properties of the materials that were assigned to the meshes. The 6061 aluminum material was a preset material already available in NX SST. The rest of the materials were created for use in this model.
The following table gives the details of the thermo-optical properties (surface finish) assigned to each mesh.
For the conductive couplings between surfaces in direct contact, for instance for surfaces that are bonded together using thermal epoxy, I applied a large conductive coupling of 10^5 W/K to signify a very high heat transfer rate. The surfaces in the model that have this direct contact coupling are:
The solar cells to the solar panels
The battery heater to the battery board
The battery board to the battery saddle
Battery saddle to power module bottom
I calculated the conductive couplings through bolted joints by applying the conduction equation (Q=[kA/L]*dT) , using the appropriate fastener area, conductivity, and length between surfaces being bolted. If a standoff was used as the fastener, the fastener area used in the equation is the ring shaped area calculated using the outer and inner diameters of the standoff. An example calculation is shown below for the coupling from the ADCS board to the module , followed by a table with the information used to calculate all conductive couplings through joints used in the model.
Q=kA/L [W/K]
Q=(163)(pi/4)[0.0047625^2-0.0028448^2 ]/(0.003175 )
Q=0.5882270172 W/K
The heat loads on the satellite that are applied in the simulations can be separated into two categories: internal heat loads and external heat loads. The internal heat loads are from heat that is produced onboard the satellite, and the external heat loads are based on the orbital parameters.
The internal heat loads in the spacecraft come from two sources: intentionally applied heat that is created by heaters, and unintentionally produced heat that comes from power dissipation by electronic components.
Battery Heater
In these simulations, the heater heat load is set to 1.5 W, which is the power that the heater on the satellite will supply. The heater heat load is applied to the top surface of the battery heater mesh in the model, and the heater controls are set to turn the heater on when the control element falls below 12 C, and turn the heater off when the control element is above 20 C, to mimic the heater control algorithm that the flight software runs. The control element for the heater control was selected as the closest element to the planned thermistor mounting location, as the battery thermistor will be used to determine whether the heater should be off or on. This heat load is applied in both the worst case hot and the worst case cold simulations, as the battery heater is always active (as in, the battery heater can always draw power) but the heater is only drawing power when the batteries are below 20 C.
Component Power Dissipation
To make a very conservative estimate of the worst case cold conditions, for the worst case cold simulation I did not apply any internal heat loads on the spacecraft from component heat production. While these conditions will never occur while the satellite is in orbit, this gives a good lower bound conservative estimate of the coldest conditions the satellite will experience.
For the worst case hot simulations, I applied heat loads to the electronic boards on the satellite for all of the major power drawing components. I applied these heat loads on the boards in the locations where these components will be mounted. The following table lists all of the major components that produce heat, which board they are located on, and their power dissipation. The size of these applied heat loads were taken from the satellite power budget.
The external heat loads on the satellite are derived from the orbital parameters that are input to the simulation, and are calculated by the software. These heat loads come from solar heating and earth albedo. The orbital parameters that I used for the orbital heating simulation objects are in the table to the right.
To determine the worst case hot and cold solar heating conditions, orbital simulations in STK were performed to determine the days in the year where the shortest and longest eclipses will occur. Based on these simulations (detailed on the power analysis page), the minimum eclipse was determined to occur on December 29th, and the maximum eclipse was determined to occur on September 19th. For the worst case hot simulation, I set the date in the orbital heating simulation object to December 29th, which resulted in a brief period (s) in Penumbra and no Umbra eclipse. For the worst case cold simulation, I set the date in the orbital heating simulation to September 19th.
For the spacecraft attitude in these simulations, I had the face of the spacecraft with the payload opening pointing to the sun, and the -x axis of the spacecraft aligned with the velocity vector of the satellite. The attitude is set in this way as this is the expected attitude that the spacecraft will be kept in by the ADCS algorithm . The attitude setting is illustrated in the image below.
The following two sections are the results of my worst case hot and worst case cold simulations, comparing the allowable minimum and maximum temperatures for each major unit on the spacecraft with the minimum and maximum temperatures reached during the simulations.
The following table gives the minimum temperatures reached by each unit during the worst case cold simulations. The third column, "Min Temp Allowed", is the minimum allowable temperature for that unit as stated in the thermal requirements (the requirement numbers are in the second column). The fourth column, "Min Temp (initial)", is the minimum temperature reached by that unit before it reaches cyclical convergence, ie. when the effects of the worst case cold initial temperature of -10C still have an impact on the resulting temperature profile. The first "Margin" column gives the margin between the minimum temperature from the simulations (before cyclical convergence) and the minimum allowable temperature from the requirements. The next column, "Min Temp (after stable), is the minimum temperature reached by each unit after cyclical convergence has been reached, ie. when the effects of the worst case cold initial temperature of -10C are no longer impacting the temperature profile of the spacecraft. The second "Margin" column gives the margin between the minimum temperature from simulations (after cyclical convergence) and the minimum allowable temperature from the requirements.
As we can see in the above data, all units have a positive margin for both the initial minimum temperatures and the minimum temperatures after stabilization.
For this simulation, the battery heater is activated periodically in order to keep the batteries warm. For each orbit, the battery heater is active for 3600s out of the 5560s orbit, with a constant power of 1.5W. This corresponds to a 65% duty cycle at worst case.
The following table gives the maximum temperatures reached by each unit during the worst case hot simulations. The third column, "Max Temp Allowed", is the maximum allowable temperature for that unit as stated in the thermal requirements (the requirement numbers are in the second column). The fourth column, "Max Temp (initial)", is the maximum temperature reached by that unit before it reaches cyclical convergence, ie. when the effects of the worst case hot initial temperature of 45C still have an impact on the resulting temperature profile. The first "Margin" column gives the margin between the maximum temperature from the simulations (before cyclical convergence) and the maximum allowable temperature from the requirements. The next column, "Max Temp (after stable), is the maximum temperature reached by each unit after cyclical convergence has been reached, ie. when the effects of the worst case hot initial temperature of 45C are no longer impacting the temperature profile of the spacecraft. The second "Margin" column gives the margin between the maximum temperature from simulations (after cyclical convergence) and the maximum allowable temperature from the requirements.
As we can see in the above data, all units have a positive margin for both the initial maximum temperatures and the maximum temperatures after stabilization.
For this simulation the external and internal heat loads from solar heating and component powers are sufficient to keep the batteries warm, and the battery heater is never activated.
When first reviewing the power module and battery mounting saddle there were some concerns about the amount of contact area between the saddle and the batteries, and whether there would be sufficient area for heat transfer. Based on my simulations the current contact area appears to be acceptable to meet thermal requirements. We are planning on using a thermal epoxy to fix the batteries in place in the saddle, so there will be very good thermal contact for the areas where there is contact. Since the only thermal path to the batteries is through the saddle they will be colder than the saddle as the spacecraft is warming up, and warmer than the saddle as the spacecraft cools down. These trends are shown in the two images to the right. The top picture shows the batteries and saddle when the spacecraft is warming up (batteries are cooler than the saddle), and the bottom picture shows the batteries and the saddle when the spacecraft is cooling down (batteries are warmer). These trends are both desirable from the standpoint of wanting to keep the temperature as steady as possible. We can see in both images that there is a minimal temperature gradient between the batteries and the flat portion of the saddle/battery heater in both heating and cooling conditions.
We can also use information and data on thermal gradients from the saddle (the thermistor location) to the batteries themselves from simulations and testing to help set limits for the thermal algorithm to control the primary heater, accounting for the temperature gradient between the batteries and the saddle. This is one of the things tested during our prototype heater test with the power module- quantifying the actual thermal gradient between the saddle and the batteries, to increase the accuracy of the thermal couplings between components in the power module (specifically the batteries, the saddle, and the heater) in the thermal simulations.
In Phase D, the updated spacecraft structure was used to rerun the simulation with the same mesh details, material and thermal properties, orbital parameters, coupling and heat loads. The results for both worst case cold and worse case hot.
As expected (based on the same inputs used) the results show only a small deviation from Phase C results with no more than a 1.8 degree difference for the initial and stable temperatures.