Image Image Image Image Image Image Image Image Image Image
Scroll to top

Top

Usage of the direct and preliminary extra-effect determination methods for diagnostic problem solving.

Usage of the direct and preliminary extra-effect determination methods for diagnostic problem solving.

| On 05, May 1998

Gregory Frenklach
gregoryf@avx.co.il

First of all let’s define what we will understand as a diagnostic problem. There are two types of the diagnostic problems.

  1. The first one is connected with finding a reason for the failure, which already had occurred and it’s REMOVAL.
  2. The second one is connected with exposure of the maximum possible future failures, which might occur, and their PREVENTION.

In order to solve diagnostic problems of the first type the RCA (Root Cause Analysis) methods are used. From TRIZ-based methods I could note ADF-1 (Anticipatory Failure Determination) and Diagnostic Problem Solving method, which was discussed in the March issue of The TRIZ Journal.

In order to solve diagnostic problem of the second type the FMEA-like methods are used, and the TRIZ-based methods ADF-2 and “Diversionary” Problem Solving Method, which was discussed in the April issue of The TRIZ Journal.

The weakest point of even TRIZ-based methods becomes clear when we deal with very complicated systems with feedback. The quantity of potentially “guilty” causes of failure grows exponentially. Moreover, because of feedback, the failures might be caused by a number of resources connected each other. It looks as if we could not solve an inventive problem (I want to remind the reader that, according to TRIZ-based methods, we transform the diagnostic problem into an inventive one by asking the question: “How could one cause the specific failure to happen?”) Thus, we deal with the extremely hard problem.

About ten years ago S. Litvin and V. Gerasimov developed the very promising Direct Extra-Effect Determination Method, and, as an amplification of this method, they created the Preliminary Extra-Effect Determination Method to use for the extremely hard inventive problems.

For example, suppose we have found a good solution for our inventive problem. This solution always is connected with some change in our system. The changed elements of the system are connected with other elements. According to Direct Extra-Effect Determination Method we have to check how their functioning is changed and indicate the positive and negative effects. The positive effects are our extra-effects, which make our solution stronger and the negative ones are our “extra-problems” which have to be solved. This process

changed elementè connected elementè element function

is repeated a number of times and the feedback paths are also considered… The result is the really strong solution.

But, if we could not find a good solution for our problem S. Litvin and V. Gerasimov suggested using the Preliminary Extra-Effect Determination Method. According to this method we assume that we ALREADY HAVE A GOOD SOLUTION of our problem (by some miracle way). Then we continue our work as in the Direct Extra-Effect Determination Method. This approach enables us to mobilize resources we had not considered, because they had appeared as result of a chain of changes. And very often because of feedback, it makes the solution clear.

Let’s return now to the diagnostic problem solving.

For example, consider a situation where we could not find the potential cause(s)/mechanism(s) of failure(s) (we could not resolve an inventive problem). According to Preliminary Extra-Effect Determination Method we have to assume that THE CAUSE(S)/MECHANISM(S) ALREADY EXIST. Then we use Direct Extra-Effect Determination Method in order to determine potential extra-effect(s) of failure, changing in the systems, feedback and so on. Such an approach very often makes clear potential cause(s)/mechanism(s) of failure(s) in extremely “hard” cases and does not take a lot of time.

I would like to note again that Preliminary Extra-Effect Determination Method can be used ONLY for solving of the diagnostic problems, which appear in the complicated systems with feedback (machines and processes). Otherwise it simply does not work.

Examples:

1. Machinery center.

Let’s determine what is the cause of a wrong measurement of tool wear. Usually tool wear is measured according to the motor current. Then the adaptive system changes regimes of the process.

The diagnostic problem of the first type is: “How one could cause wrong measurement of the tool’s wear?”

It is too hard in this problem to check all possible resources in order to find the “guilty” one… Let’s assume that we ALREADY HAVE WRONG MEASUREMENT in other words, the current of the motor is high.

Then:

  • The adaptive system change regimes of cutting;

  • It causes to work of the system under wrong conditions and with wrong regimes;

  • It causes vibrations due to under-loading and over-heat of the bearings;

  • The resistance of bearings gets higher;

  • This causes the torque to increase;

  • Which causes the motor current to increase.

Moreover:

  • Over-heat of bearings decreases viscosity of the oil;

  • The oil leaves the bearings;

  • It causes to additional over-heating…

Let’s describe the inventive situation:

During the work of the machinery center it’s bearings are over-heated. This causes the torque to increase (without the tool wear) and make the adaptive system go wrong. What can be done?

We can transform this situation into an inventive problem by different ways, for example:

  1. The bearings are overheated. What one can do? Thus our inventive problem is: Prevent the over-heating of the bearings.

  2. The adaptive system does not take into account the bearings’ over-heating. What could be done? Thus our inventive problem is: Measure the over-heating of the bearings.

Actually these problems are not inventive. They are usual engineering problems. According to Principle of the Minimal Changing of the System we choose the second problem and the engineering solution is to use a thermocouple in order to “inform” the adaptive system.

2. Sputtering machine with high vacuum.

The diagnostic problem of the second type: “How one could make vacuum to fail?”

Preliminary Extra-Effect: Assume that the vacuum already is failing.

Then:

  • The sensor gives signal and the cryogenic vacuum pump is switched on;

Then:

  • The layer of ice is increasing;

  • The o-rings of the pump are cooled;

Then:

  • The ice decreases efficiency of pump’s work;

  • The cold causes the o-rings crack;

The vacuum decreases! Air leaks through the cracked o-rings.

When the sputtering machine works a lot of time the vacuum is decreasing. In case of o-ring’s crack the specialists will discover the problem quickly, but the ice problem would not be so clear for them (the ice disappears while opening the pump).

So first let’s insert into hand-book point:

Make regeneration after certain time of work.

But this solution looks too ordinary. Let’s find better solution:

  1. Problem situation:
    After certain time of the cryogenic pump’s work the “mushroom” of the pump is coated by a layer of ice. The regeneration of the system causes to lost of time. What could be done?

  2. Type of the problem situation:
    Undesirable effect (UDE) in the existing system.

  3. UDE: Ice coating of the cryogenic pump.

  4. Element which is connected with UDE:
    The cryogenic pump.

  5. Function of the element which is connected with the UDE:
    To take molecules of air from the chamber space.

  6. Object of the function:
    Molecules of air.

  7. Direction of the problem solving:
    The UDE removal.

  8. Solving tool:
    Third group of standards.

  9. Specific tool:
    Transition into the super system. According to this standard we have to join the super system with homogeneous, heterogeneous alternative or anti-systems. Thus, if we would join together two cryogenic pumps, it would solve our problem. While one pump is under regeneration the other one works and we don’t interrupt the process.

  10. The idea development:
    Most of the time this additional pump does not work, thus we can use it for regeneration of the other sputtering machines.

Thus, the inventive idea is to use an additional “regeneration” pump for protection of the group of sputtering machines.

Conclusions:

  1. This article has presented the new direction for the diagnostic problem solving using the Preliminary Extra-Effect Determination Method which was used for inventive problem solving.

  2. The method is particularly appropriate for solving diagnostic problems which are connected with complicated systems with a lot of feedback.

  3. One can see a clear trend in failure determination methods development:

  1. FMEA like methods: “What might be wrong with the system?”

  2. AFD like methods: “How could one make the system fail?”

  3. Preliminary Extra-Effect Determination based Method: “Assume that the failure had already occurred!”

Because the Preliminary Extra-Effect Determination Method is well-suited to diagnostic problems that are connected with complicated system with a lot of feedback, it could be implemented for solving the diagnostic problems in fields of: Sociology, Stock Market Analysis, Pedagogic, Management and so on.

References:

  1. S. Litvin, V. Gerasimov Materials of Value Engineering and TRIZ course. Leningrad, 1990 (in Russian).

  2. G. Altshuller. Creativity as an Exact Science. NY. Gordon & Breach Science Publishers, 1984

  3. G. Altshuller, B. Zlotin, A. Zusman V.Filatov. Search of the New Ideas: From Inspiration to Technology. Kishinev. Karta Moldovenyaske, 1989 (in Russian)

  4. G. Frenklach. The Research (Diagnostic) Problems’ Classification. The March 1998 issue of The TRIZ Journal. https://the-trizjournal.com

  5. G. Frenklach. The “Diversionary” Method. The April 1998 issue of The TRIZ Journal. https://the-trizjournal.com