DATA ANALYSIS

PART 1. GRAPHICAL ANALYSIS AND CORRELATION ANALYSIS

1. Aim:
Determine the functional relationship between dimensionless groups using graphical analysis
and correlation analysis and interpret results.

2 Tasks:
2.1 Calculate the Reynolds number and the friction factor.
2.2 Draw a scatter plot of the friction factor versus the Reynolds number.
2.3 Calculate the Pearson correlation coefficient r.

3 Objective:
To establish the functional dependence between the dimensionless Reynolds number and
the friction factor of spheres settling in different fluids by using graphical and computing
the Pearson correlation coefficient in order to be able to apply the principles of physical
modeling and dimensional analysis to predict the terminal velocity of dust particles in air.

4 Software:
Microsoft Excel

5 Solution:
5.1 Task 1
TABLE 1
Distan Time, s Diamet Mass Velocit Volum Sphere Buoya Reynol Froude
ce, m er of of y, m/s e of density ncy, d numbe
particl particl sphere, ,kg/m^ kg/m^ numbe r, Fr
e, m e, kg m^3 3 3 r, Re

0.5 5.53 0.0031 0.0000 0.0904 1.6839 1128.2 128.28 293.39 0.2623
8 19 16 8E-08 8 11 04 23
0.5 3.9 0.0063 0.0001 0.1282 1.3408 1118.7 118.70 830.71 0.2641
5 5 05 4E-07 39 69 26
0.51 3.8 0.0063 0.0001 0.1342 1.3408 1118.7 118.70 869.62 0.2894
5 5 11 4E-07 39 94 5
0.5 1.27 0.0028 0.0000 0.3937 1.1495 3305.6 2305.6 1124.8 5.6486
38 01 5E-08 3 33 59 99
0.51 0.8 0.0063 0.0003 0.6375 1.3408 2692.3 1692.3 4130.7 6.5307
5 61 4E-07 5 47 4 13
0.5 0.76 0.0031 0.0001 0.6578 1.6839 7897.9 6897.9 2134.8 13.888
8 33 95 8E-08 7 68 01 64
0.51 0.45 0.0063 0.0010 1.1333 1.3408 7935.3 6935.3 7343.5 20.640
5 64 33 4E-07 4 4 37 28
0.51 0.4 0.0093 0.0035 1.275 4.2804 8344.9 7344.9 12164. 17.741
5 72 5E-07 1 11 54 19
0.5 0.0095 0.0035 1.02 4.5324 7880.9 6880.9 9918.9 11.139
0.51 3 72 6E-07 4 36 8 9
0.51 0.4 0.0127 0.0084 1.275 1.0726 7911.1 6911.1 16522. 13.061
86 7E-06 01 96 43
0.64 34.72 0.0031 0.0000 0.0184 1.6839 1128.2 208.28 1.0785 0.0109
8 19 33 8E-08 8 11 62 03
0.64 32.8 0.0031 0.0000 0.0195 1.6839 1128.2 208.28 1.1416 0.0122
8 19 12 8E-08 8 11 98 17
0.51 2.23 0.0063 0.0003 0.2287 1.3408 2692.3 1772.3 26.721 0.8404
5 61 4E-07 5 47 26 87
0.64 5.15 0.0015 0.0000 0.1242 2.1049 7601.0 6681.0 3.6356 0.9911
9 16 72 7E-09 5 52 97 11
0.64 4.99 0.0015 0.0000 0.1282 2.1049 7601.0 6681.0 3.7522 1.0556
9 16 57 7E-09 5 52 73 88
0.64 4.99 0.0015 0.0000 0.1282 2.1049 7601.0 6681.0 3.7522 1.0556
9 16 57 7E-09 5 52 73 88
0.64 1.94 0.0031 0.0001 0.3298 1.6839 7897.9 6977.9 19.302 3.4922
8 33 97 8E-08 7 68 93 34
0.51 1.79 0.0031 0.0001 0.2849 1.6839 7897.9 6977.9 16.671 2.6048
8 33 16 8E-08 7 68 02 4
0.51 1.17 0.0063 0.0010 0.4358 1.3408 7935.3 7015.3 50.930 3.0532
5 64 97 4E-07 4 4 26 95
0.51 0.7 0.0095 0.0035 0.7285 4.5324 7880.9 6960.9 127.75 5.6836
3 72 71 6E-07 4 36 65 23
0.09 41.76 0.0031 0.0000 0.0021 1.6839 831.36 413.63 0.0070 0.0001
8 14 55 8E-08 5 5 52 49
0.223 45 0.0063 0.0001 0.0049 1.3408 842.75 402.24 0.0323 0.0003
5 13 56 4E-07 7 31 78 95
0.05 58.81 0.0031 0.0000 0.0008 1.6839 1128.2 116.71 0.0027 2.32E-
8 19 5 8E-08 8 89 82 05
0.05 64.94 0.0031 0.0000 0.0007 1.6839 1128.2 116.71 0.0025 1.9E-
8 19 7 8E-08 8 89 19 05
0.11 22.56 0.0028 0.0000 0.0048 1.1495 3305.6 2060.6 0.0140 0.0008
38 76 5E-08 3 33 47 66
0.11 21.3 0.0028 0.0000 0.0051 1.1495 3305.6 2060.6 0.0148 0.0009
38 64 5E-08 3 33 78 72
0.23 7.91 0.0063 0.0003 0.0290 1.3408 2692.3 1447.3 0.1899 0.0135
5 61 77 4E-07 5 47 81 86
0.23 33.78 0.0015 0.0000 0.0068 2.1049 7601.0 6356.0 0.0111 0.0029
9 16 09 7E-09 5 52 39 75
0.223 30.2 0.0015 0.0000 0.0073 2.1049 7601.0 6356.0 0.0120 0.0034
9 16 84 7E-09 5 52 8 99
0.11 3.47 0.0031 0.0001 0.0317 1.6839 7897.9 6652.9 0.1037 0.0322
8 33 8E-08 7 68 23 46
0.18 4.98 0.0031 0.0001 0.0361 1.6839 7897.9 6652.9 0.1182 0.0419
8 33 45 8E-08 7 68 64 21
0.23 1.69 0.0063 0.0010 0.1360 1.3408 7935.3 6690.3 0.8891 0.2976
5 64 95 4E-07 4 4 99 34
0.28 1.82 0.0063 0.0010 0.1538 1.3408 7860.7 6615.7 1.0051 0.3803
5 54 46 4E-07 6 59 81 41
 5.2 Task2:

10000

1000
Friction factor

100

10

0.1
0.001 0.01 0.1 1 10 100 1000 10000 100000
Reynold number, Re

FIG. 1. SCATTER PLOT OF THE FRICTION FACTOR VERSUS THE REYNOLDS NUMBER.

5.3 Task 3:

TABLE 2
Buoyancy/ 1/Fr Friction logarithmi logarithmi correlatio Correl
fluid density factor c of c of n value
Reynold Friction coefficient
number, factor
Re

0.128281093 3.81209 0.65202 2.46744592 -0.1857352 -0.964621 -

3 6 0.96462
0.118703902 3.78607 0.59922 2.91945305 -0.2224073
3 9
0.118703902 3.45483 0.54543 2.93933423 -0.2632567
5
2.305632519 0.17703 0.54286 3.05109824 -0.2653067
2 7
1.692347391 0.15312 0.34465 3.61602784 -0.4626194
3 2
6.897967653 0.07200 0.66056 3.32935746 -0.1800868
1 1
6.935339678 0.04844 0.44689 3.86590531 -0.3497961
9 3
7.344910579 0.05636 0.55062 4.08509572 -0.2591445
6 4
6.880936314 0.08976 0.82151 3.996467 -0.0853822
7 9
6.911101093 0.07656 0.70373 4.21808783 -0.1525917
1 4
0.226392493 91.7176 27.6163 0.0328452 1.44116634
5
0.226392493 81.8541 24.6464 0.05755107 1.3917546
9 6
 1.926464555 1.18978 3.04846 1.42685686 0.48408184
7 9
7.262012582 1.00896 9.74510 0.56058769 0.98878635
9 1
7.262012582 0.94724 9.14898 0.57429438 0.96137298
9 6
7.262012582 0.94724 9.14898 0.57429438 0.96137298
9 6
7.584747449 0.28635 2.88861 1.28562319 0.46068947
4
7.584747449 0.38390 3.87268 1.22196209 0.58801167
1 1
7.625369215 0.32751 3.32157 1.70697586 0.52134369
5 2
7.566235124 0.17594 1.77054 2.10638286 0.24810618
4 2
0.332236935 6709.48 2964.75 -2.1517069 3.47198871
5 4
0.323086796 2534.05 1088.89 -1.4897499 3.03698595
5
0.093750126 43113.7 5375.75 -2.5556701 3.73043905
2
0.093750126 52569.9 6554.82 -2.5987312 3.81656126
7 7
1.655126522 1154.18 2540.73 -1.8524044 3.40495977
9 7
1.655126522 1028.86 2264.85 -1.8274449 3.35504078
4 7
1.162528025 73.6032 113.802 -0.7212909 2.05615217
7 6
5.10526231 336.114 2282.21 -1.9531507 3.35835652
2 5
5.10526231 285.777 1940.42 -1.917921 3.28789706
1 6
5.343749119 31.0117 220.406 -0.9841257 2.34322448
9 5
5.343749119 23.8543 169.537 -0.9271457 2.22926457
1
5.373766809 3.35983 24.0130 -0.0510012 1.380448
2 9
5.313862986 2.62921 18.5818 0.00224435 1.26908851
8 3
0.75780448 1.28365958

6 Conclusion:
In accordance with the analysis, it was finally established that there exists a very strong
negative linear relationship between the friction factor and the Reynolds number, which is
evident in the Pearson correlation coefficient of r = -0.965, which confirms the functional
dependence of the friction factor as f (Re) and consequently makes it possible to extrapolate
the results of the model to make predictions of the terminal velocity of dust particles in air.
 Laboratory No. 2
DATA ANALYSIS
PART II. LINEAR REGRESSION ANALYSIS AND LEAST SQUARES METHOD

1 Aim: get to know the linear regression analysis and the least squares method, understand
meaning of standard error S, the mean absolute percentage error (MAPE), determination
coefficient R2 and be able to interpret results.

2 Tasks:
2.1 Calculate the regression coefficients a and b using the least squares method, write the regression
equation with these coefficients and calculate values of friction factor.
2.2 Plot a graph of experimental friction factor and estimated friction factor versus the Reynolds
number.
2.3 Calculate the standard error of the regression Se, the standard error of the intercept Sa, the
standard error of the slope Sb, the mean absolute percentage error (MAPE) and determination
coefficient R2 and interpret results.

3 Software:
Microsoft Excel
4 Results:

4.5

3.5

3
(Friction factor) Y

2.5

1.5

0.5

0
-3 -2.5 -2 -1.5 -1 -0.5 0 0.5
(Reynold number, Re) X

Experimental friction factor Estimated friction factor

FIG. 1. GRAPH OF EXPERIMENTAL FRICTION FACTOR AND ESTIMATED FRICTION FACTOR VERSUS THE
REYNOLDS NUMBER

5 Conclusion:
Based on the linear regression analysis performed on the logarithmic data, the relationship
between the friction factor (ζ) and the Reynolds number (Re) was accurately modeled by the
equation friction factor (ζ) = 1.401 - 0.980 lgRe, which is confirmed by a high coefficient of
determination (R^2 = 0.985), indicating that 98.5% of the variance in the friction factor is
explained by the model, a low standard error of the estimate (Se = 0.119), and a mean absolute
percentage error (MAPE) of 3.85%, which classifies the model as highly accurate according to
Lewis's scale.
 Laboratory No. 3
LINEAR AND NONLINEAR MODEL

1 Aim of work: to get to know and learn to choose right regression equation. Be able to
interpret the obtained results.

2 Software:
Microsoft Excel

3 Solution:

4 Task 1:
Table 1.
 Coefficient
of Correlation
x y Equation task1 determination Coefficient r
R2
10 1 Linear 0,1709x-2,0327 0.8831
20 1.26 Exponential 0,5237e^0,0481x 0.9686
4,2685ln(x)-
30 1.86 Logarithmic
10,561
0.6913

40 3.31 Power 0,037x^1,2754 0.9173 0.939750874

0,0039x^2-
50 7.08 Polynomial (order 2)
0,1026x+1,614
0.9816

0,0039x^2-
60 9.18 Polynomial (order 2)
0,1026x+1,614
0.9816

10

9 f(x)
f(x) == 0.00390714285714286 x² 0.0481142693858645
0.523696386968118 exp( − 0.102614285714286xx)+ 1.614
R²
R² == 0.981621506635497
0.973105951824086
8 f(x) = 0.170885714285714 x − 2.03266666666667
R² = 0.883131705368162
7
f(x)f(x) = 4.26854518148743 x^1.27537766653804
= 0.0369564768196634 ln(x) − 10.5609936801347
6 R² R² = 0.691298410563832
= 0.857791593526637
5

0
0 10 20 30 40 50 60 70

FIG. 1. THE GRAPH BETWEEN X AND Y

5 Task 2:
TABLE 2.

Coefficient
Correlation
of
x y Equation task1 determination
Coefficient
r
R2
2 0.0245 Linear 0,0021x + 0,0506 0.8689
4 0.0337 Exponential 0,0496e0,0168x 0.7362
0.93216873
8 0.057 Logarithmic 0,0672ln(x) - 0,0613 0.88
 16 0.0855 Power 0,0152x0,6418 0.9377
32 0.1295 Polynomial (order 2) 2E-05x2 + 0,0054x + 0,0119 0.967

34 0.2 Polynomial (order 2) 2E-05x2 + 0,0054x + 0,0119 0.967

128 0.3035

0.35

0.3 f(x) = − 2.45349380311857E-05 x² + 0.00542059721646367 x + 0.0118992417219532

R² = 0.967024379545729
0.25

0.2

0.15

0.1

0.05

0
0 20 40 60 80 100 120 140

FIG. 2. THE GRAPH BETWEEN X AND Y.

6 Conclusion:
The order 2 regression model of the data (polynomial) was found to be the most suitable and
best-fitting equation of both datasets as the coefficient of determination increased to the highest
(R 2 = 0.9816 in Task 1 and R 2 = 0.967 in Task 2), which means that the order 2 regression
model explains more than 96 percent of the variance in the data and is the best representation of
the underlying functional relationships as compared to the linear, exponential, logarithmic and
power models.