Study on improving silicone losses in the process of mixing components of soldering machines

The paper presents a summary of the steps of the 6 Sigma methodology applied in the gluing section within an automotive company form Sibiu, Romania. We describe the synthesis of each step of the DMAIC model, a well-known model in the application of the 6 Sigma methodology. In the last chapter, we outline some general conclusions regarding the use of the 6 Sigma methodology presenting the benefits both within the automotive organization from Sibiu and its general use by industrial organizations.


Introduction
This study is conducted at an automotive company form Sibiu in Gluing Division.The approach is based on production processes.The purpose of this case study is to present a problem-solving project based on the 6 sigma methodology.This is a real problem faced by the organization, linked to the involvement of employees in research-innovation projects.
This company is part of a bigger multinational corporation.Its global products include: steering wheels, airbag systems, airbag fabrics, pyrotechnic capsules, safety belts, child safety chairs, electronic components and interior car sensors.The subsidiary of Sibiu has, as its main field, the manufacture of airbags for motor vehicles.
The airbag consists of the following parts: textile, sensor, electronic control unit and pyrotechnic capsule.Textile parts of the airbag are different depending on the type of airbag and model.Thus, it is distinguished: the driver's/steering wheel, passenger, knee, side and curtain airbags (Fig. 1), which in turn differ according to the model of the vehicle for which are intended.The driver's airbag is located inside the steering wheel.Head side airbag is located in the roof of the car.They can have different tubular shapes, or mounted in the door depending on application needs.
During the production of these products there were identified delays due to the silicon losses occurred during the mixing of components in the Gluing section.

Methodology
The Six Sigma methodology -a methodology initiated in the 1980's at Motorola as a challenge to achieve a reduction in the number of defective products -was used to identify the causes and to solve the problems [2].In order to achieve this effect, a thorough analysis of the causes and possibilities of correction was needed.Sigma is a term taken from statistics, used to measure the variation of a process to specifications or to another reference such as consumer requirements.The higher sigma level, the better the customer's requirements.Thus, it can be said: the probability of events being in the range [3]: -[-σ; σ] is of 68%; -[-2σ; 2σ] is of 95%; -[-3σ; 3σ] is of 99,73%; -[-6σ; 6σ] is of 99,9997% so, the probability that the event is not in the range [-6σ; 6σ] is of 3,4x10 -6 %: A six sigma improvement process comprises a series of steps, grouped by type of activity as follows [4]: Step 1. Definition -the deficiencies to be addressed will be clearly specified and the estimated improvement defined in measurable terms.It establishes a team for the project and allocates the resources and time it takes for the project to succeed.
Step 2. Measurement / Analysis -At this stage the team discovers the true causes of the deficiency.
Step 3. Improve -Establish improvements for identified causes.
Step 4. Control -the improvement team develops and implements control elements.
Step 5. Multiply the results -once the improvement team has positive results, there are two more responsibilities: to support other employees in organization with similar problems to apply what the team has learned from the improvement project, and to nominate other projects for solving in order to correct a deficiency.

Case study -applying the sigma methodology in the case of silicon losses occurring in the components mixing process
This case study was conducted in the Gluing section, starting from the production processes [5].

Defining
The project is applied in the gluing section for soldering machines.A 5W1H questionnaire (Table 1) applied to 25 people in the Gluing section was used as a tool for formulating the problem.Following the analysis of the results of the 5W1H questionnaires, the problem to be solved was formulated as follows: "Silicone losses when mixing the components".The project team consisted of project manager (Quality Engineer), Quality Team Chief, Prototype Engineer, Process Enhancement Engineer, Logistics Team Coordinator, Maintenance Engineer, Chief Laboratory Team.

Measuring range
The structure of the measurement stage consists of 7 steps for which specific actions have been taken.The key milestones for the masonry phase are: customer voice, setting measurement aspects, validating measurement systems, detecting influence factors, measuring planning, visualizing graphical measurements and performance of the process.

Customer voice
In the measurement phase, the expectations, requirements and specifications of the internal clients are summarized in Table 2.

Fitted parts according to the standard
The proportion of component A and B is 1: 1 Resistance to silicon at break> 175N >175N 10%

Reduce silicone scrapping due to expiration Proportional use of components A and B
Compliance with the date of validity of components A and B Using FIFO of silicone barrels 10%

Establishing aspects of measurement
In order to better understand the change process, the measurement factors have been prioritized (Table 3)

Validation of measurement systems
Validation of measuring systems was done with a specialized "Measurement System Analysis" system shown in Fig. 2.

Detection of influence factors
In order to detect the possible causes of the problem, the Ishikawa diagram was used.The main cause was "The Stock Difference Between Components A and B Generated by Ratio Variation".The mixing process is influenced by the following parameters: D1-speed on the route; D2-starting speed; D3-speed of absorption.

Planning the measurements
Leakage velocity may be influenced by viscosity (different viscosity between the two components).The influence of viscosity on the mixing process can be expressed by the difference between rations due to the slower or faster leakage of components A and B through the pipe.The monitoring of the mixing process is presented in Tables 4 and 5.

Planning the measurements
The mixing process is influenced by the following parameters: D1-speed on the route; D2-starting speed; D3-speed of absorption.
Leakage velocity may be influenced by viscosity (different viscosity between the two components).The influence of viscosity on the mixing process can be expressed by the difference between rations due to the slower or faster leakage of components A and B through the pipe.The monitoring of the mixing process is presented in Tables 4 and 5.A range of values for the three machine parameters for silicone type A and type B has been established and their influence on the ratio by two methods has been monitored.Ratio by method 1 -the classical method used in the line is allowed to flow 40 seconds silicone type A and then type B, after which weigh.Fig. 3.The weighing process of the type A component [1] Ratio by Method 2 consists of depositing A and B type silicon on the panel, according to a route corresponding to each model (BMW F10).

Fig. 4. Silicone component type A deposition process [1]
The values obtained by changing the parameters on the ratio through the two test methods for components A and B are presented in Table 6.Following the modification of the parameters, the values are obtained which are highlighted in the graphs of Fig. 5 and Fig. 6, taking the and base ratio in percent (difference between components A and B in grams).

Find and prioritize the main causes that impact on performance
For identification of the root cause the following questionnaire was made (Table 8).
Due to the fact that the ratio is higher than the maximum admissible 5%, the consumption of the two components A and B is different.
Table 9 shows the difference in Kg between January and May.
Following the calculations, the difference between consumption (formula 1) and ration (formula 2) was double.

Ration =[(A -B) / (A + B)]*100
(2) The different perception of consumption and ration has made for values of rationality considered good (difference to 2.5% versus maximum specification of 5%), in fact to be too high for the proposed objective, because the calculation formulas for consumption and ration give totally different results.Consumption 1% different from ration 1%. 43

Updating project benefits (based on root cause analysis)
Following the implementation of the optimal parameters on head 5, the average ration of 0.28% was obtained, comparing it with the average of the ration in April (2.71%) there was a decrease of the ration by 10.33%, but comparing with the change in machine parameters is a drop of only 6%.By making an analogy for head 6 (which rate in April was 2.17%) to bring it to a value of 0.22%, (which means a 8.11% reduction in the rate) should be we modify machine parameters not by 8%, but by 4%.

Generating ideas for solutions
The evaluation criteria for the alternatives previously developed are: total cost, impact on the problem, cost / benefit relationship, resilience / impact on change, implementation time, uncertainty about efficacy.
The main solutions that were found following a brainstorming session (Table 10):

Cause
Alternative 1 Alternative 2 a.Not respecting the ration a.1.Reducing the specified range from a maximum of 5% to 2.5% a.2.Adjustment of rations in process to 2.5% and their validation by silicon breaking strength > 175N b.Deregulation of parameters on machine 3 head 6 b.1.Adjusting the parameters of the machine 3 the head 6 according to the optimal parameters set for the head 5 b.2.Following the ratios obtained from the alignments of the head 6 with optimal parameters from head 5 c.Lossless monitoring c.1.Drafting Loss Sheets and communicating these losses to the logistics department

Prioritize solutions
In order to evaluate the improvement alternatives against these criteria, the team used as a quality tool the alternative selection matrix (Table 11) where it used the following notations: 3 -Very favorable impact, 2 -Favorable medium impact, 1 -Poor impact .Adjustment of process ratios to 2.5% and their validation by silicon breaking strength> 175Nprioritization of ideas was made on the basis of the large difference between the result obtained by the calculation formula of consumption and ration Adjusting the machine parameters 3 the head 6 according to the optimal parameters set for the head 5 -on the assumption that it is the same machine and the same type of motors it was considered that optimal parameters obtained for M3 C5 can be applied to C6.

Designing solutions
Due to the change in machine parameters, there were little differences between settings 3 and 10 where the track speed was higher for silicone A and large differences between settings 1 and 12 where track speed was higher for silicon B. For machine 3 head 5 the parameters giving the best result for the 0.22 ration were the ones corresponding to the setting 3. The next step was to modify the current parameters (setting 2) with the parameters at setting 3 on both the head 5 and the head 6.
From the top data it can see that setting 3 is the optimal one because a lower ration is obtained by modifying parameter D21.

Validation of solutions
After changing the machine parameters to setting 3 to validate the 0.22% rate on both the head 5 and 6, the ration was made and the silicone resistance to breakage was tested by 2 methods.The ration on the head 5 confirms the initial result, on the head 6 the ratio exceeds the limit of 5% admitted.Results are confirmed by the silicone tear strength test by: Method 1 -Rapid drying of the parts (in the oven 30 minutes at 80° C) and Method 2 (classic) after 24 hours under normal temperature conditions.

Measuring the Impact of Improvement
As a result of the improvement of parameters D21, D31, only a reduction of the ration from 8,33% to 7,69% during the period January-May 2015 was obtained on the machine 3 head 5, compared to the target value (3,5%) representing improvement of 13.25%.

Control
Based on the results of the DOE parameter analysis, the improvement was achieved with the modification of parameters D21 and D31 only (parameters D21 and D31 have a major influence on ration, D22, D32 have little influence, and D23, D33 insignificant infinity).So extending to all silicone machines can be applied in a simpler way just by doing 4 attempts.

Conclusions
As a result of the improvement of the only on the machine 3, the head 5 achieved an improvement of 13.25% which represents 573 kg of silicon B, and 1 kg of silicon costs 12.54 €.Financial benefits 7185 € / 5months Which means a benefit of 1437 € / 1 month.
If the action applied to the machine 3 the head 5 would extend with the same results on M1C1 machines (machine 1 head 1), M1C2 (machine 1 head 2), M2C3 (machine 2 head 2), M2C4 (machine 2 head 4), and M3C6 (machine 3, head 6) would have a financial benefit of 7185 eur.
Six Sigma is a tool that, if used correctly, can identify the key areas of a business with the advantage that all of the improvements measured and obtained through this technique are directly transformed with financial results.6 sigma is a philosophy that relies on good resourcesaving and good behavior practices that are geared to the continuous improvement of current work and life so that everything we do and everything we live is more enjoyable, better, more beautiful, cheaper and simpler.

Cause1 3 3 2 14 5 MATEC 1
The alternative of improvement Selection criteria Total a b c d e f a.Not respecting ration a.1 Reducing the specified range from a maximum of 5% to 2.5% 1 1 3 1 2 2 10 a.2 Adjustment of rations in process to 2.5% and their validation by silicon breaking strength> 175N 2 3 Web of Conferences 184, 04021 (2018) https://doi.org/10.1051/matecconf/201818404021Annual Session of Scientific Papers IMT ORADEA 2018 Adjusting the parameters of the machine 3 the head 6 according to the optimal parameters set for the head 5

Table 2 .
Customer voice

Table 3 .
Prioritization of measurement factors Y

Table 4 .
Monitoring of the mixing process during January 24-February 03 2017 on first exchange.

Table 5 .
Monitoring of the mixing process during January 24 -February 06 2017 on second exchange

Table 8 .
The 5 WHY questionnaire

Table 9 .
Comparative analysis between the two methods