Dough Thickness for kurtoskalacs

1
Dough Thickness for kurtoskalacs

• By
• Timothy Ruggles and Heath Whittier

2
Chimney Sweets - Kurtoskalacs

• Kurtoskalacs is a centuries old dessert from
  Transylvania, a type of sweet bread wrapped
  around a wooden dowel and baked quickly over an
  open fire or in a rotisserie oven.
• Over the summer I worked at the Provo Farmers
  Market baking them. The process has been
  unchanged for centuries and is difficult to get
  right.
• The challenge to making the perfect kurtoskalacs
  is making the dough the right thickness so they
  cook quickly and the sugar on the outside
  caramelizes while the dough inside stays soft.

3
The Problem Predicting the required dough
thickness

• Sugar carmelizes at 160 C
• Dough is cooked at 93 C
• Using these two temperatures for the boundaries
  on each side of the dough we set up an equation
  to predict the perfect thickness.
• Because the dough is wrapped around a dowel we
  used the Approximate analytical solution for
  transient conduction through an infinite
  cylinder.

4
Adapting the Model

• This analytical solution was not meant for a
  non-homogenous cylinder. Because the dough and
  the dowel have differing properties, this model
  will not necessarily be accurate.
• The conduction coefficient was taken to be an
  average between dough (apprx. 0.4 W/mK) and wood
  (0.16 W/mK), 0.28 W/mK.
• We set the time for the inside to finish cooking
  and the time for the outside sugar to caramelize
  equal to each other.

5
Solution and results

• The equation simplified to the following
• We found ? from table 5.1 using Bihr/k, and
  assumed it did not change with t.
• Estimated h20, k.28, r.0381m
• Bi2.72 ?1.7
• Using Mathcad we then solved for t
• Which gives us t 0.25595 inches

6
Conclusions

• Our model predicted the ideal thickness for a
  crispy outside and a soft inside to be about a
  quarter of an inch.
• This value is close to common practice, but still
  too small. Our model needs further refinement to
  accurately predict baking parameters.

7
Recommendations

• The inaccuracy of our solution may be due in part
  to the inadequacy of the model and poorly
  estimated properties and parameters.
• Oven temperature and convective coefficient were
  estimated because the oven was unavailable. If
  they were measured, more precise results could be
  obtained.
• The non-homogenous nature of the problem lends
  itself to a finite element analysis. Although
  our model was good for an initial guess, FEA
  would yield better results.

8
Applications

• Right now, Chimney Sweets, the company I worked
  for, is developing new ovens. A working model
  for the temperature of a kurtoskalacs as it bakes
  could help in oven design, setting target oven
  temperatures and convective coefficients in order
  to preserve the traditional taste of the
  kurtoskalacs.