Turducken Thermal Design and Analysis

Timeline:
Fall 2017

Tools used:
ANSYS Workbench, Altair Hypermesh, SolidWorks, MATLAB

Background

The final project for Cooper Union’s Introduction to Computer Aided Engineering (CAE) course involved teams of 4 students analyzing the cooking process of a turducken using a convection oven and 2 resistive heating skewers. The turducken consists of 3 layers: an outer layer of turkey, a middle layer of a homogeneous mixture of duck and chicken, and an inner layer of team-selected stuffing. The objective was to minimize the cooking time required to fully cook the turducken (minimum temperature of 185°F) while also minimizing the percentage of burnt turducken (temperature exceeding 265°F). We performed a transient thermal analysis on the turducken using ANSYS Workbench and validated our results with simplified hand calculations.

Turducken geometry: turkey layer in orange, duck/chicken layer in gray, and stuffing layer in blue.

Each team was provided with the same turducken geometry, temperature criteria, and design limitations for the skewers and oven. The full assignment with all applicable rules can be found here.

Setup and Mesh

Our team designed simple pointed skewers, placing them through the thickest parts of the turducken as these areas have the highest thermal capacitance. Using the 2006 ASHRAE Refrigeration Handbook, we determined the relevant thermal conductivity, density, and specific heat for each material as a function of temperature, weighting homogeneous mixtures gravimetrically for specific heat and volumetrically for thermal conductivity and density. We additionally accounted for the changes in these properties due to water evaporation and material burning.

Turducken mesh.

Turducken with skewers side and rear views.

To reduce our mesh and solve time, we exploited geometric and thermal symmetry, splitting the turducken along its vertical midplane. All convection coefficients and emissive properties were approximated based on similar studies in standard convection ovens, as physical experimentation and CFD methods were outside the scope of this analysis.

Turducken boundary conditions.

Using Altair Hypermesh, we sliced the stuffing, ducken, wing, and skewers to allow for mapped hexahedral meshing where possible. Because the turkey is not as easily mappable, we opted to use a hexcore mesh primarily for mesh space efficiency. Where hexahedral elements could not be used, we used smaller tetrahedral elements, notably at locations where we expected higher temperature gradients (e.g., around the skewer). The mesh had a total element count of approximately 495,000. All elements were element type SOLID70.

Validation

To validate our model results, we created an equivalent transient thermal RC circuit to simulate the rise in temperature at various nodes across the turducken. This model treated the turducken as a set of concentric spheres, lumping the resistive and capacitive effects for each layer. The skewers were treated as current sources that were toggled at specific intervals in the analysis. The resulting coupled ODEs were assembled in MATLAB and are solved concurrently using Forward Euler numerical methods.

Turducken thermal circuit.

Hand calculation results indicate that the skewers should remain on for approximately 4.25 hours before being shut off to minimize the percentage of burnt turducken while drastically reducing cooking time.

Results

Our model results showed close agreement with the hand calculations (maximum ~20% difference). Results indicated that, if the skewers were left on for 4.25 hours and if the turducken was left in the oven for 6.25 hours, the turducken was only 39% burnt. Compared to a similar analysis without the skewers, the turducken with skewers was 20% less burnt and needed about 6 fewer hours in the oven.

Turducken heating animation.

Hand calculation heating results Hand calculation results compared to analysis results: solid lines represent ANSYS simulation results, and dashed lines represent hand calculation results.