ASHRAE Journal - August 2019 - 8

LETTERS

Automatic
Commercial Ice
Makers
Without a doubt, the May Journal
article "Simulation of Efficiency
Improvements to an Automatic
Commercial Ice Maker" is of great
interest both from a scientific and
technical point of view. Simulation
of refrigeration equipment with
detailed analysis (more than 80
variables were used) with subsequent experimental tests on a
commercially available unit is very
important.
The authors outlined the specific
test conditions and presented a
comparison between theory and
experiment. The small absolute
difference in percentages between
experiment and modeling (the
authors called it "model's accuracy
within 5%") allowed them to draw
far-reaching conclusions regarding
the possible improvement of the ice
maker performance.
As far as I know, only 20% of publications of the International Journal
of Refrigeration for 2012 - 2018 are
at this level of research, which combines theory, simulation, testing
industrial ice generator, and comparing theoretical and experimental
data.
To my great chagrin, I have to add
a spoonful of tar to the barrel with
honey. I find it difficult to agree with
some of the authors' statements.
First of all, the model accuracy
claimed by the authors is not the
model uncertainty and cannot be
used to confirm the validity of the
theoretical model. If we assume that
three experiments were carried out
under different operating conditions
8

ASHRAE JOURNAL

ashrae.org

(Table 1), then it is incredibly not
enough to calculate the mismatch
between computational predictions
(CP) and experimental results (ER),
| CP - ER |. Its definition can use the
mean absolute deviation.
If experimental uncertainty EU >
|CP - ER|, the value of the proposed
model is insignificant, and it is very
risky to put it into practice. Thus, the
stated well correlation between the
experimental and calculated data
does not guarantee that the choice
of the model structure will be sufficiently complete.
Even if we assume that all (N=80)
variables are calculated and measured with very high accuracy, for
example 1% (it is very difficult to
believe), then the total uncertainty
achieved of key performance criteria
Δ, for example energy consumption
in a day, will be equal to:
Δ = K·(∑Ni=1 Δi2)1/2 =
1.2·(80·0.012)1/2 = 1.2·0.089 = 0.107 =
10.7%
where K is a correction factor equal
to 1.2 for the most important parameters of the devices and monitoring
the characteristics of the finished
product.
This value covers much of the
declared model's accuracy.
A "statistical significance" between
the results of theory and experimental data is not sufficient evidence
of the correctness of the chosen
model. A necessary condition for the
accuracy of the model is, above all,
the smallness of the calculated total
uncertainty of the objective function compared with the mismatch
between theory and experiment.
This fundamental truth has not had
an important place in the refrigeration

A U G U S T 2 0 19

literature, but must be the subject of considerable debate.
I hope the authors will find a possibility to confirm their assumptions.
Boris M. Menin, Ph.D., Member ASHRAE,
Beer-Sheva , Israel

The Authors Respond
The authors acknowledge the concerns of Dr. Boris Menin. We tried to
be careful about our wording with
respect to the comparison between
the simulation model and the limited experimental results. We did
not intend to make broader claims
about model uncertainty.
The model input parameters used
were obtained from physical dimensions or correlations taken from
existing literature. Model parameters were not adjusted to improve
model correlation. Our intention
was to illustrate the correlation
between the simulation model with
the experimental data that was
available. We identified a model
"accuracy for these measures,"
meaning operating points available.
We understand that far more
experimental data is required for
a comprehensive statement about
model accuracy. Still, the promising
correlation at those operating conditions provided us confidence to evaluate design improvements at those
same operating conditions. We
apologize for the misunderstanding
and our unclear selection of words.
Haithem Murgham, Ph.D.,
David Myszka, P.E., Dayton, Ohio;
Kyaw Wynn, Associate Member ASHRAE, Sydney, Ohio

Letters to the Editor should
be no more than 250 words and
must relate to an article published in ASHRAE Journal.
Please send your letters to
journaleditor@ashrae.org.

https://www.ashrae.org/

ASHRAE Journal - August 2019

Table of Contents for the Digital Edition of ASHRAE Journal - August 2019