1.
Failure Rate
R(t) = e(-λt)
0.5 = e(-λ100)
Ln(0.5) -λ100
λ= 0.006931 Failure per hours
2. Probability 500 hours Failure
R(t) = e(-λt)
R(t) = e(-0.00693 x 500)
R(t) = e(-3.465)
R(t) = 0.031273
2. MTTF and Survival probability
MTTF = 1/λ
MTTF = 144.2695 hours
R(t) = e(-λt)
R(t) = e(-0.00693 x 144.27)
R(t) = 37% chance to survive
50 Hours
R(t) = 1 - θ [(ln(t) - μ') / δ]
1 - θ [(ln(50) - 20) / 10]
1-θ -1.61 => Using Z Table
1- 0.0537
R(t) = 0.9463
200 Hours
R(t) = 1 - θ [(ln(t) - μ') / δ]
1 - θ [(ln(200) - 20) / 10]
1-θ -1.47 => Using Z Table
1- 0.0708
R(t) = 0.9292
ln[1/(1-
Failure No, j TTF (Days) Ln (TTF) Median Rank Median
Rank)]
1 120 4.79 0.1296 0.1388
2 520 6.25 0.3148 0.3781
3 700 6.55 0.5000 0.6931
4 1180 7.07 0.6852 1.1558
5 6320 8.75 0.8704 2.0431
Distribution Normal Distribution Lognormal Distribution Exponential
Paremeter Value Paremeter Value Paremeter
Mean 1768.00 Log mean 6.68 lamda
Stdev 2572.96 Log Stdev 1.43 MTTF
F(t), Distribution Absolut [Dist - Median
Normal Lognormal Exponential Weibull Normal
0.2609 0.0931 0.0418 0.1503 0.13
0.3138 0.3822 0.1688 0.3646 0.00
0.3390 0.4632 0.2204 0.4277 0.16
0.4096 0.6071 0.3427 0.5523 0.28
0.9616 0.9254 0.8943 0.9253 0.09
Normal 0.28
Lognormal 0.08 => The lowest KS Value ~ Lognormal is the Best Distribu
KS Value
Exponential 0.34
Weibull 0.13
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.930318331595
R Square 0.865492198101
Adjusted R Sq 0.615492198101
Standard Error 0.454855833474
Observations 5
ANOVA
df SS MS F
Regression 1 5.3250441244 5.32504412443 25.7380519
Residual 4 0.827575317 0.20689382925
Total 5 6.1526194414
Coefficients Standard Error t Stat P-value
Intercept 0 #N/A #N/A #N/A
X Variable 1 0.000355607611 7.009438E-05 5.07326836921 0.00711413
ln(ln[1/(1-
Median Rank)])
-1.9745
-0.9727
-0.3665
0.1448
0.7145
Distribution Exponential Distribution Weibull
Value ParemeterValue
0.0003556 Beta 0.69829543621
2812.09 Intercept -5.15789495676
Eta 1613.89847363
Absolut [Dist - Median]
Lognormal ExponentiaWeibull
0.04 0.09 0.02
0.07 0.15 0.05
0.04 0.28 0.07
0.08 0.34 0.13
0.05 0.02 0.05
Value ~ Lognormal is the Best Distribution
Significance F
0.01478998869
Lower 95% Upper 95% Lower 95.0% Upper 95.0%
#N/A #N/A #N/A #N/A
0.00016099441 0.00055 0.00016099441 0.00055
Old Design New Design
ln(ln[1/(1- ln[1/(1- ln(ln[1/(1-
TTF ln TTF ln[1/(1-Median F(t), Weibull R(t), Weibull ln TTF F(t), Weibull
No. (Weeks) (Weeks) Median Rank Rank)] Median Distribution Distribution No. TTF (Weeks) (Weeks) Median Rank Median Median Distribution
Rank)]) Rank)] Rank)])
1 2 0.6931 0.0246 0.0250 -3.6906 0.0490 0.9510 1 15 2.7081 0.0673 0.0697 -2.6638 0.0612
2 2 0.6931 0.0599 0.0617 -2.7851 0.0490 0.9510 2 32 3.4657 0.1635 0.1785 -1.7233 0.1569
3 3 1.0986 0.0951 0.0999 -2.3036 0.0713 0.9287 3 61 4.1109 0.2596 0.3006 -1.2020 0.3282
4 4 1.3863 0.1303 0.1396 -1.9691 0.0928 0.9072 4 67 4.2047 0.3558 0.4397 -0.8217 0.3623
5 6 1.7918 0.1655 0.1809 -1.7097 0.1338 0.8662 5 75 4.3175 0.4519 0.6013 -0.5086 0.4065
6 9 2.1972 0.2007 0.2240 -1.4960 0.1908 0.8092 6 116 4.7536 0.5481 0.7942 -0.2304 0.6032
7 11 2.3979 0.2359 0.2691 -1.3128 0.2262 0.7738 7 148 4.9972 0.6442 1.0335 0.0329 0.7198
8 17 2.8332 0.2711 0.3163 -1.1512 0.3222 0.6778 8 178 5.1818 0.7404 1.3486 0.2990 0.8022
9 19 2.9444 0.3063 0.3658 -1.0057 0.3512 0.6488 9 181 5.1985 0.8365 1.8112 0.5940 0.8092
10 21 3.0445 0.3415 0.4179 -0.8726 0.3788 0.6212 10 183 5.2095 0.9327 2.6985 0.9927 0.8138
11 23 3.1355 0.3768 0.4728 -0.7490 0.4051 0.5949
12 28 3.3322 0.4120 0.5310 -0.6330 0.4658 0.5342 Weibull Distribution Answer = New design prove that has better realibility than the old one
13 33 3.4965 0.4472 0.5927 -0.5230 0.5199 0.4801 Paremeter Value
14 34 3.5264 0.4824 0.6585 -0.4177 0.5300 0.4700 Beta 1.31153277434
15 34 3.5264 0.5176 0.7290 -0.3161 0.5300 0.4700 Intercept -6.3131907651
16 37 3.6109 0.5528 0.8048 -0.2172 0.5589 0.4411 Eta 123.17392852
17 38 3.6376 0.5880 0.8868 -0.1201 0.5682 0.4318
18 40 3.6889 0.6232 0.9761 -0.0241 0.5860 0.4140
19 45 3.8067 0.6585 1.0743 0.0716 0.6274 0.3726 Old Design vs New Design - Reliability
20 55 4.0073 0.6937 1.1831 0.1681 0.6976 0.3024
1.0000
21 56 4.0254 0.7289 1.3052 0.2663 0.7038 0.2962
22 57 4.0431 0.7641 1.4443 0.3676 0.7099 0.2901 0.9000
23 67 4.2047 0.7993 1.6059 0.4737 0.7642 0.2358 0.8000
24 76 4.3307 0.8345 1.7988 0.5871 0.8040 0.1960 0.7000
25 90 4.4998 0.8697 2.0381 0.7120 0.8528 0.1472
Reliability R(t)
0.6000
26 115 4.7449 0.9049 2.3531 0.8557 0.9113 0.0887
27 126 4.8363 0.9401 2.8158 1.0352 0.9289 0.0711 0.5000
28 197 5.2832 0.9754 3.7031 1.3092 0.9826 0.0174 0.4000
0.3000
Weibull Distribution
0.2000
Paremeter Value
Beta 0.956715 0.1000
Intercept -3.654845 0.0000
0 20 40 60 80 100 120 140 160 180 200
Eta 45.61353
TTF (Weeks)
Old Design New Design
New design is significantly better as it has a higher reliability according to the chart
R(t), Weibull
Distribution
0.9388
0.8431
0.6718
0.6377
0.5935
0.3968
0.2802
0.1978
0.1908
0.1862
etter realibility than the old one
ln[1/(1- ln(ln[1/(1-
TTF Median
No. Ln (TTF) Median Median
(Days) Rank Rank)] Rank)])
1 7 1.95 0.0343 0.0349 -3.3548 Answer
2 12 2.48 0.0833 0.0870 -2.4417 Normal 0.26854
3 15 2.71 0.1324 0.1420 -1.9521 Lognormal 0.16590
KS Value
4 20 3.00 0.1814 0.2001 -1.6088 Exponential 0.18176
5 23 3.14 0.2304 0.2619 -1.3399 Weibull 0.15164
6 100 4.61 0.2794 0.3277 -1.1157
7 105 4.65 0.3284 0.3981 -0.9210 MTTF = 285.3102019
8 110 4.70 0.3775 0.4739 -0.7467
9 120 4.79 0.4265 0.5559 -0.5871
10 130 4.87 0.4755 0.6453 -0.4381
11 140 4.94 0.5245 0.7434 -0.2965
12 150 5.01 0.5735 0.8522 -0.1599
13 160 5.08 0.6225 0.9743 -0.0260
14 500 6.21 0.6716 1.1134 0.1074
15 520 6.25 0.7206 1.2751 0.2430
16 530 6.27 0.7696 1.4680 0.3839
17 550 6.31 0.8186 1.7072 0.5349
18 570 6.35 0.8676 2.0223 0.7042
19 590 6.38 0.9167 2.4849 0.9102
20 610 6.41 0.9657 3.3722 1.2156
Distribution NormalDistribution Lognormal
Distribution ExponentialDistribution Weibull
ParemeterValue ParemeterValue ParemeterValue ParemeterValue
Mean 248.10 Log mean 4.81 lamda 0.0036327 Beta 0.776312
Stdev 235.23 Log Stdev 1.45 MTTF 275.28 Intercept -4.274735
Eta 246.2789
F(t), Distribution Absolut [Dist - Median]
Normal LognormalExponentiaWeibull Normal Lognormal ExponentiaWeibull
0.1527 0.0246 0.0251 0.0611 0.12 0.01 0.01 0.03
0.1578 0.0553 0.0427 0.0913 0.07 0.03 0.04 0.01
0.1609 0.0746 0.0530 0.1077 0.03 0.06 0.08 0.02
0.1661 0.1067 0.0701 0.1327 0.02 0.07 0.11 0.05
0.1693 0.1254 0.0802 0.1468 0.06 0.10 0.15 0.08
0.2645 0.4453 0.3046 0.3915 0.01 0.17 0.03 0.11
0.2715 0.4586 0.3171 0.4030 0.06 0.13 0.01 0.07
0.2786 0.4713 0.3294 0.4143 0.10 0.09 0.05 0.04
0.2930 0.4952 0.3533 0.4358 0.13 0.07 0.07 0.01
0.3078 0.5171 0.3764 0.4561 0.17 0.04 0.10 0.02
0.3229 0.5374 0.3986 0.4753 0.20 0.01 0.13 0.05
0.3383 0.5562 0.4201 0.4936 0.24 0.02 0.15 0.08
0.3540 0.5737 0.4408 0.5110 0.27 0.05 0.18 0.11
0.8579 0.8338 0.8374 0.8232 0.19 0.16 0.17 0.15
0.8761 0.8404 0.8488 0.8324 0.16 0.12 0.13 0.11
0.8846 0.8436 0.8542 0.8368 0.12 0.07 0.08 0.07
0.9003 0.8496 0.8644 0.8452 0.08 0.03 0.05 0.03
0.9144 0.8553 0.8739 0.8531 0.05 0.01 0.01 0.01
0.9270 0.8606 0.8827 0.8606 0.01 0.06 0.03 0.06
0.9380 0.8656 0.8909 0.8676 0.03 0.10 0.07 0.10
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.953146
R Square 0.908487
Adjusted R 0.855856
Standard E 0.3996
Observatio 20
ANOVA
df SS MS F Significance F
Regression 1 30.11904 30.11904 188.6211 5.5834E-11
Residual 19 3.033922 0.15968
Total 20 33.15296
Coefficients
Standard Error t Stat P-value Lower 95% Upper 95%Lower 95.0%Upper 95.0%
Intercept 0 #N/A #N/A #N/A #N/A #N/A #N/A #N/A
X Variable 0.003633 0.000265 13.73394 2.569E-11 0.0030791 0.004186 0.003079 0.0041862948
LAMBDA for Exponential Distribution
=> The lowest KS Value ~ Weibull is the Best Distribution
Time Between Cumm. TBF
Failure Number Mann Test
Failure (Days) (Days) 2. Failure 180 days after
1 410 410 1 Y=
2 47 457 7 Y=
3 207 664 2 Y=
4 112 776 5 Y=
5 997 1773 0
6 391 2164 0
7 162 2326 2
8 186 2512 0
9 3 2515 1
10 168 2683 0
Total 18
1. Graphical Test, Mann, Laplace Test
Graphical Test
Exp Trend Test Linear Trend Test
12 12
10
Failure Number
10
Failure Number
8 f(x) = 1.41226720872326 exp( 0.000715752392404762 x ) 8 f(x) = 0.00307659946361093 x + 0.491296073241398
6 R² = 0.84608973227507 6 R² = 0.916341842664823
4 4
2 2
0 0
0 500 1000 1500 2000 2500 3000 0 500 1000 1500 2000 2500 3000
Cumm TBF (days) Cumm TBF (days)
Trend R2 Value
Exponential Trend 0.9430
Linear Trend 0.9163
Natural log Trend 0.8807
Exponential Upward concave is more fit to the trend, NHPP is selected for futher analysis
Mann Test (Confidence Level Assumption 90%)
M value
Degradation Current Improvement
14 18 31
Mann Current Value is in between degradation and Improvement Renewal Process assumption is valid
Laplace Test
Failure Truncated (90% confidence level)
a= 0
b= 2683
Tj= 13597
n= 9
L= 0.65567881857
No trend is detected, RP assumption is valid.
ure 180 days after
1.4123e0.0007x
1.4123e0.0007(2683+180)
1.4123e2.0041
10.47844 Failure
rend Test Ln Trend Test
12
10
Failure Number
8
9946361093 x + 0.491296073241398 f(x) = 3.72701644576995 ln(x) − 21.2446510832584
2664823 6 R² = 0.88069447677252
4
2
0
1500 2000 2500 3000 0 500 1000 1500 2000 2500 3000
m TBF (days) Cumm TBF (days)
Cumm Time
Failure Number Time Between Between Failure Mann Test
Failure (hours)
(hours)
1 1030 1030 3
2 2800 3830 0
3 970 4800 3
4 2150 6950 0
5 640 7590 2
6 1600 9190 0
7 1090 10280 0
Total 8
Graphical Trend Test
Exp Trend Test Linear Trend Test
8 8
f(x) = 0.915119989289755 exp( 0.000209435813289306 x )
Failure Number
Failure Number
6 R² = 0.972349137885182 6 f(x) = 0.000661467210400256 x − 0.126610439739884
R² = 0.974719182182663
4 4
2 2
0 0
0 2000 4000 6000 8000 10000 0 2000 4000 6000 8000 10000
Cumm TBF (hours) Cumm TBF (hours)
Trend R2 Value
Exponential Trend 0.9677
Linear Trend 0.9747
Natural log Trend 0.7934
Linear is more fit to the trend, HPP constatnt Failure Rate is selected for futher analysis
Mann Test (Confidence Level Assumption 90%)
M value
Degradation Current Improvement
5 8 16
Mann Current Value is in between degradation and Improvement Renewal Process assumption is valid
Laplace Test
Assume Failure Truncated (90% confidence level)
a= 0
b= 10280
Tj= 33390
n= 6
L= 0.35080200234
No trend is detected, RP assumption is valid.
2. 2 years after renewal failures
ln[1/(1-
Failure No, j TTF (Days) Median Rank Median
Rank)]
1 1030 0.0946 0.0994
2 2800 0.2297 0.2610
3 970 0.3649 0.4539
4 2150 0.5000 0.6931
5 640 0.6351 1.0082
6 1600 0.7703 1.4709
7 1090 0.9054 2.3582
SUMMARY OUTPUT
Regression Statistics
Multiple R 0.626315529
R Square 0.39227114186
Adjusted R Square 0.22560447519
Standard Error 0.98121165425
Observations 7
ANOVA
df SS MS F Significance F
Regression 1 3.728663112733 3.728663113 3.872824 0.106202
Residual 6 5.776657862563 0.96277631
Total 7 9.505320975296
CoefficientsStandard Error t Stat P-value Lower 95%Upper 95%Lower 95.0%
Intercept 0 #N/A #N/A #N/A #N/A #N/A #N/A
X Variable 1 0.00044746947 0.000227378573 1.967949163 0.096625 -0.000109 0.001004 -0.000109
Exponential Distribution
lamda 0.00044746947
Expected number of failure for next 2 years (17520hours)
T2= 27800 = 2 years
T1= 10280 = last recorded cumm TTF
Expected number of failure
7.83966515847057 ~7 failures
3. MTBF
MTBF = 2234.79
d Test Ln Trend Test
8
Failure Number
0256 x − 0.126610439739884 6
f(x) = 2.43841571663509 ln(x) − 16.8342474439618
4 R² = 0.793411907326717
2
0
6000 8000 10000 0 2000 4000 6000 8000 10000
F (hours) Cumm TBF (hours)
Upper 95.0%
#N/A
0.001004
Cumm Time
Failure Number Time Between Between Failure Mann Test ln (Ti)
Failure (Days)
(Days)
1 193 193 2 5.2627
2 126 319 2 5.7652
3 18 337 4 5.8201
4 61 398 3 5.9865
5 243 641 0 6.4630
6 115 756 1 6.6280
7 211 967 0 0.0000
Total 12 35.9255
1. Graphical Trend Test, Mann, Laplace
Exp Trend Test Linear Trend Test
8 8
7 f(x) = 1.10029674512851 exp( 0.00217560287427612 x ) 7
R² = 0.783414907394513 f(x) = 0.00751510228821061 x + 0.123280805324496
Failure Number
6
Failure Number
6 R² = 0.939387786026327
5 5
4 4
3 3
2 2
1 1
0 0
100 200 300 400 500 600 700 800 900 1000 100 200 300 400 500 600 700 800 900 1000
Cumm TBF (days) Cumm TBF (days)
1 1
F
0 0
100 200 300 400 500 600 700 800 900 1000 100 200 300 400 500 600 700 800 900 1000
Cumm TBF (days) Cumm TBF (days)
Trend R2 Value
Exponential T 0.8314
Linear Trend 0.9394
Natural log Tr 0.9687
Ln is more fit to the trend, NHPP constatnt Failure Rate is selected for futher analysis
Mann Test (assume 90% confidence level)
M value
Degradation Current Improvement
5 12 16
Mann Current Value is in between degradation and Improvement Renewal Process assumption is valid
Laplace Test
Assume Failure Truncated (90% confidence level)
a= 0
b= 967
Tj= 2644
n= 6
L= -0.3759
No trend is detected, RP assumption is valid.
2. 2 years after
As Bad as Old (NHPP) (Failure Truncated)
T= 967
r= 6
beta= 1.12788
lamda= 0.00258
T1= 967
T2= 1697 365 x 2 days = 730 days
Expected number of failure after 2 years
5.314687462
~ 6 failures
3. MTBF Last day failure for the next 3 years
T1= 967
T2= 2062
MTBF= 135.26650584 days
end Test Ln Trend Test
8
8821061 x + 0.123280805324496 7
f(x) = 3.76535400509406 ln(x) − 19.0222810606571
Failure Number
6327 6
5 R² = 0.968660446901247
4
3
2
1
0
0 600 700 800 900 1000 100 200 300 400 500 600 700 800 900 1000
m TBF (days) Cumm TBF (days)
1
F
0
0 600 700 800 900 1000 100 200 300 400 500 600 700 800 900 1000
m TBF (days) Cumm TBF (days)
