Rand Simulation Solutions Blog

Can a Turkey be Cooked with Microwaves?

If you could realistically microwave a turkey this Thanksgiving, would you?

Taking inspiration from last Thanksgiving’s Supersonic Turkey Cooking (STC) using Ansys Fluent by CFD Specialist Alex Lefebvre, I developed a handheld microwave emitter and simulated the feasibility of using it to cook this year’s Thanksgiving turkey. A traditional microwave oven utilizes a magnetron to convert power into microwaves (the common frequency of microwave ovens is 2.4GHz). These microwaves travel from the magnetron to the oven chamber through a waveguide (think air ducts for RF signals). To create a handheld version of a microwave, I utilized the Antenna Design Toolkit within Ansys HFSS. Using this toolkit, I selected an elliptical horn antenna because it includes a waveguide section and had the directionality that would ensure I wasn’t going to cook my hand alongside the turkey. All I needed to do to generate my antenna model was to then input the desired operating frequency of 2.4GHz and let the HFSS toolkit take care of the rest.

Fig 1: Ansys HFSS Antenna Toolkit Setup

With the waveguide/horn antenna complete, the rest of the HFSS model was created. The simulation was performed using 500W of power applied to the wave port on the antenna input. In Figure 2, one can observe the fields propagating from the emitter reaching field strengths more than 5000 V/m! This is seriously something you should not try at home!

Fig 2: Ansys HFSS RF Simulation (500W)

When microwaves are absorbed by the turkey, the energy contained within the RF fields are converted into losses in the form of heat. Using the simulated electromagnetic losses absorbed by the turkey, Ansys Icepak was used to solve the steady-state temperature of the turkey due to the microwave emitter. The results of our 500W simulation (Figure 3) show that the turkey breast will reach a temperature over 500 degrees Fahrenheit! A bit overdone for my taste. We also see that the underside of the turkey is still raw!

Fig 3: Ansys Icepak Results (500W)

It is apparent that the direct radiation of the microwaves is causing an uneven distribution of fields which in turn causes these extremes in the turkey’s temperature. A metallic vessel was then created to prevent direct RF radiation to the turkey while also trapping the microwaves around the turkey. The power level was also reduced from 500W to 300W. In Figure 4, one can see that the fields deflect off the top and continue to bounce around within the vessel. Similar to the internal walls of a traditional microwave, this effectively acts as an oven containing the microwaves (our source of heat).

Fig 4: Ansys HFSS RF Simulation (w/Vessel @ 300W)

Repeating our linked simulation in Icepak shows that we have achieved the target internal temperature of 165 degrees Fahrenheit for the safe consumption of turkey. However, it can be observed in Figure 5 that the turkey breast still has excessive heat. It is expected that further tuning of the vessel’s deflector is necessary for those that desire extra-extra-crispy turkey this Thanksgiving. Stay safe, do not take apart your microwave and try to do this (seriously), and have a great Thanksgiving!

 

Fig 5: Ansys Icepak Results (w/Vessel @ 300W)

 

Will I get wetter if I walk through the rain or if I run?

Answers to Enduring Questions

We have all been in this situation and, at least briefly, asked ourselves one of the enduring questions, “Will I get wetter if I walk through the rain or if I run?”  I pondered this question, and the possibility of CFD answering it while watching it rain outside my office more than 10 years ago but never got around to trying to answer it until now.  Part of the delay in answering this question was with the software I was using at the time.  To use CFD to answer this question the software must show a solid body, a human body, in this case, moving through a fluid volume. 

Figure 1:  Human body, with region for overset mesh, in an air domain

It also must be able to show particles of water, or at least particles with the density of water, moving through the fluid volume. 

Figure 2: Air Domain with water particle injection from the top surface (particle size shown at 25x)

The CFD package I was using 10 years ago had limitations.  It could solve transient solid body motion.  It could also solve discrete particle motion through an air volume.  I could not, however, separate the particles which hit the human body from those which hit the ground.  The move to Ansys Fluent made this possible and added several opportunities for improving the analysis. 

Setting up the analysis requires defining the solid body motion speed and parameters to quantify rainfall, both how they enter the domain and how they exit.  To define solid motion, the person moves through the domain at three different speeds, 0.5, 2, and 5 m/s relating to walking, jogging, and sprinting.  The rain is defined as starting from a particle injection surface with randomized starting points and staggered positions on the top surface of the domain.  Particles are injected moving downward at 9 m/s downward as if they have reached terminal velocity.  The diameter of particles varies from 0.2mm to 2mm, which is a common range of raindrop diameters.  The total mass of particles is held constant at 1 Kg/s where anything over 2 Kg/s is “flash flood” range so 1 Kg/s is a steady rain of small to medium size raindrops.  

Both the bottom surface of the domain and the surface of the human body are defined as escape surfaces as if the particles are immediately absorbed by whichever surface they contact.  The graph below shows the distance the person traveled vs the mass of accumulated water that hits the person.  Walking yields the highest mass of absorbed water, above 0.18 kg.  Jogging, 4 times walking speed, reduces wet mass below 0.04 kg and absorbed water at sprinting speed, 10 times walking speed, is even further reduced.  Moving faster through the rain reduces how wet you will get. 

Figure 3: Surface Velocity through a Zero-Depth Entrance and Exit Region

 

So, does this answer the enduring question?  Significant simplifications were employed including eliminating effects of water-shedding, saturation, and splashing from footfalls.  The answer also may change under different rain conditions or with wind added.  The results also don’t account for the increased risk of slipping and falling at higher movement speeds.  This is one answer, under one set of conditions with several significant assumptions.  It is a significant result that provides opportunities for future study.    

Ansys Fluent is a powerful and flexible tool that can help solve many enduring questions.  If you have an application where not all your questions are being answered, contact Rand Simulation to discuss how CFD can help.

Power Integrity Analysis Using Ansys SIwave

Printed circuit boards (PCBs) are becoming more and more complex. Having multiple clocks, multiple power supplies, and many data channels on a single PCB has become more of a norm. As the complexity of PCBs increase as do their demands on their power delivery networks (PDN). These complex systems require that the voltage on the PDN is stable. The margin for voltage ripple has become smaller and smaller as designers need to ensure the reference voltage is stable to have confidence in their device performance.

When investigating voltage ripple we need to make sure we are discussing all that entails power integrity within a device. That means the many inductive loops, capacitive planes, and at times the complex resistance paths need to be considered. That is why it is imperative that designers use a simulation tool to simulate their PDN and perform power integrity (PI) analysis. Ansys provides their users a streamlined workflow through Ansys SIwave to perform PI analysis with a step-by-step workflow wizard shown below.

Fig 1. Ansys SIwave workflow wizard

Ansys SIwave allows users to import many different board designs including ODB++, IPC-2581, and DXF. This allows users of popular PCB designer tools such as Altium, Cadence, or Mentor Graphics to quickly bring in their PCB into Ansys tools. Once imported, pad stacks and stackups can be verified and components can be mapped through the inclusion of S-parameter or spice models or through the internal component library to Ansys SIwave. After these verifications and component mappings, users can then begin to set up their first simulation.

The DCIR simulation allows users to assign voltage sources and current sinks to visualize the voltage, current, and power on their traces, layers, and vias. As well as providing power mapping and verification of proper voltage and current supply to critical components. Figure two depicts an example of a voltage result from Ansys SIwave. In addition to a single-phase VRM, multi-phase VRM systems can be simulated within the DCIR analysis. One of the advantages of performing the DCIR analysis in Ansys SIwave is that the user can use the resulting data as a source for a CFD thermal simulation within Ansys Icepak. So the user can take plane, trace, and via level information for their thermal simulations.

Figure 2. Voltage results from DCIR Simulation

Within the PI analysis, users will extract the SYZ parameters for their PCBs. Within the tool, there is a smooth workflow for assigning ports to VRMs, ICs, and decoupling cap positions to get a full system extraction of the SYZ-parameters.

Users can then take the extractions into the PI Advisor Wizard to optimize the decoupling capacitors on their board. Required impedance characteristics can be set over frequency, as well as the number of capacitors to optimize. Figure three depicts the PI Advisor Wizard.

Figure 3. PI Advisor in Ansys SIwave

Once simulated, the PI advisor will give several schemes that outline the capacitors and their values as well as an estimated price. In addition, the tool will show the new impedance for the user-chosen scheme. Figure four demonstrates the optimized results for one of those schemes.

With the new PDN being optimized and simulated, the z-parameter results can be taken into the Ansys circuit tool for a transient analysis. This allows users to fully include spice and component models for their PDN as well as the newly extracted impedance. The transient simulation can be performed, and the resulting voltage ripple can be examined to verify that it meets the component manufacturer’s spec.

Ansys SIwave and Ansys Circuit can be used to perform frequency and transient domain simulation on PDNs. Using the PI wizard, users can easily optimize their decoupling capacitors and their placements. As well as perform a DCIR analysis to ensure proper voltage and current are getting to their ICs. In addition, the DCIR results can be used as sources for a CFD thermal simulation on the user’s PCB. With Ansys tools, users can simulate complex PDNs and optimize their designs to ensure they are meeting the intricate and at times difficult requirements of their ICs.