1 Application of TD for TCM.ppt

Presented By

Author 1: Dr. Koilakuntla MadduletyAssistant Professor

Affiliation: Operations Management Group National Institute of Industrial

EngineeringMumbai – 400087; India,

E Mail: [email protected],Mobile: +91-9969326007

1

6th International Working Conference ’’TOTAL QUALITY MANAGEMENT - ADVANCED AND INTELLIGENT APPROACHES’’, With

Second Special Conference ’’Manufuture in Serbia 2011’’, 7th – 11th June, 2011, Belgrade, SERBIA.

A CASE STUDY: APPLICATION OF TAGUCHI APPROACH FOR PROCESS OPTIMIZATION OF TABLET COMPRESSION MACHINES

OF HLL LIFECARE LIMITED, INDIA.

Index Cum Methodology

Introduction Problem Statement Objectives Identification of Response Variables Determination of factors and

Levels Design for Experimentation Conducting Experiments Analyses Interpretations and arriving

optimal setting Validation of Optimal Setting: Conclusions

2



Introduction

Quality – Fitness for use Any product not meeting the requirements

leads to cost of poor quality Occurrence of rejections are either from

common or special causes Robust design refers to designing a

product or a process in a way that it has minimal sensitivity to the external noise factors

3



Introduction

Taguchi techniques It is an Offline Quality assurance technique It is a fractional factorial design of

experiments Robust designs in different phases

System design Parameter design Tolerance design

Signal to Noise ratios Smaller the better Larger the better Nominal the better

4



Introduction5

Brief Profile of HLL Lifecare Limited: HLL Life-care Limited Kanagala Facility, Belgaum,

India: The Kanagala plant in Belgaum, Karnataka, India commenced its operation with production of condoms in 1985 using Japanese technology. This unit underwent diversification in 1992 with the tabletting manufacturing facility for birth control pills - Mala-D/N and the formulation and tabletting of Saheli (Centchroman) the indigenous, non-steroidal once-a-week pill. The tabletting of Emergency Contraceptive pills was started in 2003 (7)



Problem Statement

Around 15% batches were reworked in every month due to non attainment of either required Compression-Strength which is to be Grater-than 1 Kg/Cm2, (Higher the Better Case) or Friability Less-than 1% (Lower the Better Case) that are output quality characteristics of Tablets.

6



Objectives of Study

Acquiring knowledge of deployment of

Taguchi Approach for solving Problem Deploying the Taguchi Approach at

Problem area systematically in 10 steps as above

Ensuring Rework reduction from 15% to 5% by optimum setting of input parameters

7



Identification of Response Variables8



The Brainstorming Technique was used by involving all the concerned employees and executives and decided to optimize two Quality Characteristics of Tablets that are:

1. Compression-Strength (Hardness), & 2. Friability

for reducing rework from 15% to 5%

Determination of Factors & Levels

9



Identification of the factors and factor levels that are influencing above two response Variables:

After application of Brainstorming technique with all the concerned employees and executives and after establishing case and effect relations between input- parameters and output-parameters of compression process the most significant four process parameters are identified as control parameters along with levels and shown in table below:

Determination of Factors & Levels

10



Sl. No.

Name of the FactorNotation

Unit of Measure

No. of Factors Levels

Level Values

1 2 3

1 Percentage of Moisture in granules

M % 3 2 2.3 2.6

2 Speed of the Machine S RPM 3 25 28 31

3 Release Pressure RP Level 3 2 2.5 3

4 Weight Dozer WD Level 3 1 2 3

11



Sl. No. M S RP WD Note1 2 25 2 1

2 2 28 2.5 2 Name of the FactorNotation

3 2 31 3 3 Percentage of Moisture in granules

M4 2.3 25 2.5 35 2.3 28 3 1 Speed of the Machine S6 2.3 31 2 2 Release Pressure RP

7 2.6 25 3 2Weight Dozer WD

8 2.6 28 2 39 2.6 31 2.5 1

L9 Taguchi Design for Experimentation is arrived by Minitab software

Design for Experimentation

Conducting Experiments

12



:

As per above design 9 experiments with two replications for each run and collected the two values that are Friability and Hardness of Tablets for each replication of each run and posted in Minitab worksheet shown below:

Experimentation13



Sl. No. M S RP WD F1 F2 H1 H21 2 25 2 1 0.4095 0.4125 2.35 2.352 2 28 2.5 2 0.24 0.2325 3.1 3.073 2 31 3 3 0.6765 0.693 1.3 1.2284 2.3 25 2.5 3 0.465 0.48 2.14 2.085 2.3 28 3 1 0.615 0.6075 1.54 1.576 2.3 31 2 2 0.585 0.5775 1.66 1.697 2.6 25 3 2 0.765 0.7575 0.94 0.978 2.6 28 2 3 0.51 0.495 1.96 2.0299 2.6 31 2.5 1 0.51 0.525 1.96 1.9

Note1 First Replication’s Friability Value F12 Second Replication’s Friability Value F23 First Replication’s Hardness Value H14 Second Replication’s Hardness Value H2

Experimental Output

Analysis

Analyses of data of Tablets’ for selected Quality Characteristics Friability and Hardness

Optimization by ANOVA and Taguchi

Approach with the help of Minitab Software and

Interpretation of Analyses and selection of the optimum levels are as follows:

14



Analysis

Analysis of Variance for Friability, using Adjusted SS for Tests

Source DF Seq SS Adj SS Adj MS F P M 2 0.072495 0.072495 0.036248 552.34 0.000 S 2 0.065345 0.065345 0.032672 497.86 0.000 RP 2 0.239791 0.239791 0.119896 1826.98 0.000 WD 2 0.004996 0.004996 0.002498 38.06 0.000 Error 9 0.000591 0.000591 0.000066 Total 17 0.383218

S = 0.00810093 R-Sq = 99.85% R-Sq(adj) = 99.71%

15



Analysis16



Analysis of Variance for Hardness, using Adjusted SS for Tests

Source DF Seq SS Adj SS Adj MS F PM 2 1.19780 1.19780 0.59890 550.46 0.00S 2 1.08344 1.08344 0.54172 497.90 0.00RP 2 3.88520 3.88520 1.94260 1785.48 0.00WD 2 0.07988 0.07988 0.03994 36.71 0.00Error 9 0.00979 0.00979 0.00109Total 17 6.25610

S = 0.0329848 R-Sq = 99.84% R-Sq(adj) = 99.70%

Analysis of Friability (Means)

17



2.62.32.0

0.7

0.6

0.5

0.4312825

3.02.52.0

0.7

0.6

0.5

0.4321

M

Mean o

f M

eans

S

RP WD

Main Effects Plot for MeansData Means

Analysis of Friability (SN Ratios)

18



2.62.32.0

8

7

6

5

4

312825

3.02.52.0

8

7

6

5

4

321

M

Mean o

f SN ratios

S

RP WD

Main Effects Plot for SN ratiosData Means

Signal-to-noise: Smaller is better

Analysis of Hardness (Menas)

19



2.62.32.0

2.4

2.1

1.8

1.5

1.2312825

3.02.52.0

2.4

2.1

1.8

1.5

1.2321

M

Mean o

f Means

S

RP WD

Main Effects Plot for MeansData Means

Analysis of Hardness (SN Ratios)

20



2.62.32.0

8

7

6

5

4

312825

3.02.52.0

8

7

6

5

4

321

M

Mean o

f SN ratios

S

RP WD

Main Effects Plot for SN ratiosData Means

Signal-to-noise: Smaller is better

Analysis of Mean Delta Values for Friability

21



Response Table for Means of Friability

Level M S RP WD1 0.4440 0.5483 0.4983 0.51332 0.5550 0.4500 0.4087 0.52633 0.5938 0.5945 0.6857 0.5532Delta 0.1498 0.1445 0.2770 0.0400Rank 2 3 1 4

(Smaller is better)

Analysis of ‘SN Ratio’ Delta Values for Friability

22



Taguchi Analysis: F1, F2 versus M, S, RP, WD Response Table for Signal to Noise Ratios for Friability


Analysis of Mean Delta Values for Hardness

23



Response Table for Means of Hardness (Kg per Cm2)(Larger is better)


Analysis of ‘SN Ratio’ Delta Values for Hardness

24



Taguchi Analysis: H1, H2 versus M, S, RP, WD Response Table for Signal to Noise Ratios of Hardness


Interpretations and arriving optimal setting

25



Based on the ‘General Liner Model ANOVA’ developed by Minitab software twice (shown below), once Friability as response variable and second time Hardness as response variable for investigating significance effect of four input variables M, S, RP & WD on Friability and Hardness and concluded that all the four input variable M, S, RP & WD (all the p-values are 0.00 i.e less than 0.05 and linear model R-square value is more than 99%) are significantly effecting the both the responses- Friability and Hardness

The optimal setting has been arrived after developing and observing Friability Main Effect plots for means, Hardness Main Effect plots for means, Friability Main Effect plots for SN Ratios, Hardness Main Effect plots for SN Ratios, and by considering all the delta values for Means and SN Ratios


26



Name of the FactorNotation

Optimal Level

Percentage of Moisture in granules M 2%

Speed of the Machine S 28 RPM

Release Pressure RP 2.5 Level

Weight Dozer WD 1st Level

Optimal setting is arrived after considering main effect plots and delta values

The Optimal Setting is 2% moisture in granules; 28 RPM Speed; Release pressure at 2.5 Level; and Weight Dozer at 1st Level gives for Minimum Friability and Maximum Hardness.

27



The predicted value for optimal setting has been arrived by Minitab Software for both the parameters Friability and Hardness are as follows

•The predicted Friability for above optimal setting is 0.223% with 0.006 Standard Deviation

•The predicted Hardness for above optimal setting is 3.125 Kg per Cm2 with 0.0212 Standard Deviation


Validation of Optimal Setting:

28



50 batches have been produced with above optimal setting and checked the Friability and found the average as 0.23% and highest value is as 0.26% and average Hardness is 3.0 Kg/cm2 with lowest hardness in 50 batches is 2.7 Kg/cm2 and found no batch is rejected or reworked, all the batches had been passed for next operation that is film Coating and Blister packing.

CONCLUSION

The Robust Parameter Design through Taguchi Approach is shown a breakthrough Improvement in compression process parameter optimization i.e average Hardness Improved from ‘1.5 Kg/Cm2 with 0.5 standard deviation’ to ‘3.0 Kg/Cm2 with standard deviation of 0.02’.

and average Friability reduced from 0.6 to 0.23 at the same time the Friability standard deviation reduced from 0.5 to 0.01, which were breakthrough improvement in both the parameters.

29



REFERENCES

Ross, Philip J., ‘ Taguchi techniques for Quality Engineering’ Prentice hall, 1989.

Klien, I.E. ‘Improving the yield in production of thin RF integrated circuits’, International Journal of Microelectronics Vol. 13, Issue 3, pg 12-14, December 1996

Der Ho Wu and Hsin Hua Chen, ‘Application of Taguchi robust design method to SAW mass sensing device’, IEEE Transactions on Ultrasonic, Ferroelectrics and Frequency Control, Volume 52, Issue 12, Page(s): 2403 – 2410, ,( December 2005)

30



REFERENCES

Palanikumar, K. ‘Cutting Parameters Optimization for Surface Roughness in Machining of GFRP Composites using Taguchi’s Method’, Sathyabama Deemed University.

Montgomery, Douglas C., ‘Design and Analysis of Experiments’, 5th edition, Wiley edition, 2006,

Bagchi, Tapan P., ‘Taguchi methods explained – Practical steps to Robust design’, Eastern Economy edition, 1993.

http://www.lifecarehll.com/adminRight.aspx?id=21&par_id=21

31



THANK YOUAny Questions

Please?

32

6th International Working Conference ’’TOTAL QUALITY MANAGEMENT - ADVANCED AND INTELLIGENT APPROACHES’’, With Second Special Conference ’’Manufuture in Serbia 2011’’, 7th – 11th June, 2011,

Belgrade, SERBIA.

E Mail: [email protected]

1 Application of TD for TCM.ppt

Documents

special conference manufuture

intelligent approaches

introduction quality

special causesrobust

limited kanagala facility

external noise factors

noise ratiossmaller

kanagala plant