Dr Tansu KÜÇÜKÖNCÜ — RETRACTED (March 2015) article which says “results of my neural network perfectly match theory, while my experimental results do NOT” : Bekir Karlık, Feb 2015

Dr Tansu KÜÇÜKÖNCÜ —
RETRACTED (March 2015) article which says
“results of my neural network perfectly match theory,
while my experimental results do NOT” : Bekir Karlık, Feb 2015

Report : “PLAGIARIZED” – “Self PLAGIARIZED” – “DUPLICATION” article which has “IRREPRODUCIBLE Results”

Bekir Karlik

[RETRACTED article]
[53 % PLAGIARIZED article]
—————————————
Betul Sarac, Bekir Karlik, Umit Uncu, Teoman Ayhan,
“Neural Network Methodology for Modeling Heat Transfer in Wake Flow”,
ASME Journal of Heat Transfer, Feb 2015, Vol. 137, pp. 022201-1..10

 

A. PLAGIARISM – SELF PLAGIARISM – DUPLICATION :

A.1. Plagiarized Text :

Turnitin “similarity report” :
“similarity ratio” is 53 %.

Sources of “copy-paste” given by Turnitin “similarity report” :
[1]. 20% match (publications)
Sarac, B..
“Heat transfer from a short plate behind an obstacle in tandem”,
Energy Conversion and Management, 201109
(Betul Sarac, Hussain Al-Madani, Teoman Ayhan,
“Heat transfer from a short plate behind an obstacle in tandem”,
Energy Conversion and Management 52 (2011) 3147–3158)

[2]. 11% match (publications)
International Journal of Numerical Methods for Heat & Fluid Flow,
Volume 17, Issue 8 (2007-10-21)
(Betul Ayhan-Sarac, Bekir Karlık, Tulin Bali, Teoman Ayhan,
“Neural network methodology for heat transfer enhancement data”,
International Journal of Numerical Methods for Heat & Fluid Flow,
Vol. 17 No. 8, 2007, pp. 788-798)

[3]. 3% match : NOT IMPORTANT

[4]. 3% match (publications)
Daloglu, A..
“Heat transfer from a cylinder in the wake flow”,
International Communications in Heat and Mass Transfer, 200005

[5]. 2% match (publications)
Industrial Lubrication and Tribology, Volume 64, Issue 5 (2012-08-18)

6 x 1% match
41 x <1% match

“Table 1” is completely copy-pasted from “Table 1” of [1],
but not completely colored in Turnitin “similarity report”.

“Nomenclature” and “Dimensionless Terms” are completely copy-pasted
from “Nomenclature” of [1],
but not completely colored in Turnitin “similarity report”.

These mean that copy-pasted text ratio is more than 53 % .

A.2. Plagiarized Figures :

Turnitin “similarity report” “similarity ratio” does not include “copy-pasted” figures.

Sources of “copy-pasted” figures :
It has 11 titles (Fig.1. .. Fig.11.) for 36 figures.
13 of figures are copy-pasted from the same source which is also
the source of the 20 % copy-pasted text :
[1]. 20% match (publications)
Sarac, B..
“Heat transfer from a short plate behind an obstacle in tandem”,
Energy Conversion and Management, 201109
(Betul Sarac, Hussain Al-Madani, Teoman Ayhan,
“Heat transfer from a short plate behind an obstacle in tandem”,
Energy Conversion and Management 52 (2011) 3147–3158)

13 figures which are copy-pasted from [1] without citation :
Fig. 3. a, b, c.               (= Fig. 1. a, b, c of [1])
Fig. 4.                           (= Fig. 2. a, b of [1])
Fig. 5.                           (= Fig. 3. of [1])
Fig. 6.                           (= Fig. 4. of [1])
Fig. 7.                           (= Fig. 5. of [1])
Fig. 8. a, b, c, d, e, f.     (= Fig. 6. a, b, f, d, e, c. of [1])

13 figures = 54.17 % (= 13 / 24) of total figures.

<– o –>

B. IRREPRODUCIBLE Results :

B.1. NOT GIVEN details and Contradictions with another paper :

Some contradictions with another paper :
[2]. 11% match (publications)
International Journal of Numerical Methods for Heat & Fluid Flow,
Volume 17, Issue 8 (2007-10-21)
(Betul Ayhan-Sarac, Bekir Karlık, Tulin Bali, Teoman Ayhan,
“Neural network methodology for heat transfer enhancement data”,
International Journal of Numerical Methods for Heat & Fluid Flow,
Vol. 17 No. 8, 2007, pp. 788-798)

Both have almost the same title :
“Neural Network Methodology for Modeling Heat Transfer in Wake Flow”
“Neural network methodology for heat transfer enhancement data”

For [2] :
402 experimental measurements are collected.
About 225 of those are used as training data for neural networks,
the rest is used for testing.

For “subject paper” :
no of experimental measurements : NOT GIVEN.
no of experimental measurements used for training : NOT GIVEN.
no of experimental measurements used for testing : NOT GIVEN.

B.2. NOT GIVEN details and self Contradictions :
while experimental results do not match with theoretical results so good,
neural network results almost perfectly match with theoretical results

Some contradictions in results :
no of experimental measurements : NOT GIVEN.
no of experimental measurements used for training : NOT GIVEN.
no of experimental measurements used for testing : NOT GIVEN.
but
while experimental results do not match with theoretical results so good
according to Fig. 9-a, 10-a, and 11-a,
–> neural network results almost perfectly match with theoretical results
–> according to Fig. 9-b, 10-b, and 11-b.

According to Fig. 10-b. : NuNNet < 0.8
but
according to Fig. 10-c. : NuNNet < 1.0

B.3. NOT GIVEN details and self Contradictions :

There is no reference to “Figs. 10. a, b, c”, and “Figs. 11. b”.

There is 1 reference to “Figs. 11. a, c”, but at an unrelated place which is
copy-pasted from [1] (see. “Conclusions”) without citation and
original text in [1] does not have such a reference :
“The effect of the plate on the rate of heat transfer decreases as
the aspect ratio increases,
although this effect is not consistent at all times, depending on
the characteristics of the wake streets”

There is 1 reference to “Fig. 9”, but at an unrelated place which is
copy-pasted from [1] (see. pg. 3156, right column, paragraph 2) without citation and
original text in [1] does not have such a reference to “Fig. 11″of [1]
which is similar to “Fig. 9.a” :
“From the general appearance of Figs. 6–8 it can be seen the Nusselt number variations
trough the flow direction have similar behaviours, but the magnitudes of the average
Nusselt numbers differs due to the plate orientation angles and the unstable wake generation behind the obstacles.”

B.4. NOT GIVEN details and self Contradictions :

Figs. 8. a, b, c, d, e, f. (which were copy-pasted from
Fig. 6. a, b, f, d, e, c. of [1] without citation) have hints about parameters used.

“Figs. 9. a, b, c”, “Figs. 10. a, b, c”, “Figs. 11. a, b, c” have no hint about parameters.

Contradiction between
Fig. 6. (which were copy-pasted from Fig. 4. of [1]),
Fig. 7. (which were copy-pasted from  Fig. 5. of [1]),
Figs. 8. a, b, c, d, e, f. (which were copy-pasted from
Fig. 6. a, b, f, d, e, c. of [1] without citation), and
“Figs. 9. a, b, c”, “Figs. 10. a, b, c”, “Figs. 11. a, b, c” :
In Fig. 6., Fig. 7., Figs. 8. a, b, c, d, e, f.
Nusselt numbers are greater than 10 and less than 80.
But in “Figs. 9. a, b, c”, “Figs. 10. a, b, c”, and “Figs. 11. a, b, c”
Nusselt numbers are greater than 0 and less than 1.1.

In [1] there is a figure similar to “Fig. 9. a”, “Fig. 10. a”, and “Fig. 11. a” ,
in it Nusselt numbers are greater than 10 and less than 110.

B.5. NOT GIVEN details and Contradictions with another work :

Contradiction with [1] :
20 % copy-pastes from [1] are not as 1 block, but as several pieces
including manipulations to sentences which causes shifts in the meaning.
For example, according to [1] this is not true :
(see “Abstract” and “pg. 8, right column, paragraph 3”)
“Results from all configurations studied show that the location of
maximum heat transfer behind the obstacle was found to be
directly dependent on the Reynolds number and
–> aspect ratio.”
According to original sentences in [1] (see “Conclusions”) :
“– The results clearly describe wake phenomena behind the an obstacle
as highly effected by the aspect ratio (D/H) and flow Reynolds number.
– The location of maximum heat transfer behind the obstacle in the wake region
was found to be directly dependent on Reynolds number.
– The effect of plate angle on the rate of heat transfer decreases as the
–> aspect ratio increases, although
–> this effect is not consistent at all times
–> depending on the characteristics of the wake street.”