Help me in Evaluation of Xfema package

[Rajesh Goyal]

Qu-I:
I am evaluating the Xfema software (Using MIL-STD), Now when I am giving Data to `Add Failure GUI', I am not very clear about `Failure Probability' and `Probability of Loss', when I am comparing my data with TASK-101 and TASK-102. In TASK-102 columns are `Failure Effect Prob (Beta)' and `Failure Mode Ratio (alpha)'.
(a) I am using `Failure Effect Prob' of TASK-102 as `Probability of Loss' in Xfema Add Failure GUI and `Failure Mode Ratio' of TASK-102 in `Ratio of Unreliability' - IS IT OK?
(b) What is `Failure Probability'? in relation to TASK-102.
Qu-II:
When I am adding/editing any Item and Clicking on `Reliability..' button, The GUI pop-up requires Distribution and Parameters. I am selecting `Exponential', It asks for Lambda, Now I am having "Failure Rate Lamda(p) in 10^6" ( e.g. 1.237) what value of lambda is to be given in Xfmea.
Qu-III:
Can I calculate the Mission Reliability also using the system Hierarchy? and if so then how we can enter the number of components
AN EARLY REPLY WILL BE APPRECIATED AS MY EVALUATION LIC IS EXPIRING SOON.
RAJESH

[Rajesh Goyal]

Can you also explain the functional limitations of Xfema evaluation version

[tarik/lance]

When using the Mil-Std 1629 to perform your analysis in Xfmea, you will need to decide if you will be using quantitative or qualitative criticality analysis. The Failure Mode Ratio of Unreliability and Probability of Loss are required for quantitative criticality analysis The Failure Probability and Severity Class are required for qualitative criticality analysis.

The lambda is simply the failure rate. There are no units associated with it. Please remain consistent throughout your analysis. For example, if your measures are in terms of hours, convert all of your entries to hours and all of the outputs will be in terms of hours. I would also like to remind you that you can show the mean life instead of the lambda if you prefer. This is done by selecting User Setup from the File menu. Then click on the Other tab, and change the setting under Exponential Parameter as required before clicking on OK.

Regarding your last question, I am not sure what you mean by calculating the mission reliability using the system hierarchy. Do you mean that you will be defining reliabilities for your systems, sub-systems, and components, and would like to use these values to calculate system reliability? If so, then you can do this by exporting your Xfmea project as XML and then importing it into BlockSim to perform your system analysis. The only drawback there is that this assumes a series connection (no complex interconnections) between your components.

The evaluation is a time-limited version that gives you 45 days to demo the software. Currently, there are only minor restrictions on this evaluation; it is very similar to the full version.

[Rajesh Goyal]

Tkanks Tarik and Lance,
Please refer to my Qu-II and your reply in para-2 regarding lambda.
I am keeping my inputs as number of failures per 10^6 Hrs. But when I am comparing my results with one of my old analysis (which I am using to evaluate the software), I am not getting results from Xfmea as it should be. The `Unreliability' column for all sub-system and componets is not calculated correctly, but rest of the coeficient are ok (with 10^6). The moment I am changing value of Lembda (as 1/MTBF in Hrs), the `Unreliability' part (Xfmea Column) is shown correctly but other coefficient are not as expected (as per MIL STD Format i.e TASK 102). In case you are not clear about what I want to convey, I can send a part of my data for one sub system to check at your end.

[Rajesh Goyal]

Dear Lance Larson
Thanks for your reply. I know analysis in the software complies with Mil-Std 1629. Today was the last day for my evaluation copy, and I was going through the Report Generation Feature of the software. What I have noticed is that, in the Xfmea you are giving:
Mode Criticality = Unreliability X Mode Ratio X Probability of Loss ....(1)
whereas as per Mil Std 1629, Task-102 Format
Mode Criticality = Failure Mode Ratio X Probability of Loss X Failure Rate X Time ....(2)
Now, report generated by Xfmea also gives the same definition in Header Row of Report (quantitative criticality analysis) but calculations are based on Equation (1) above.
I am not able to understand this. To my understanding you can't deviate by using `Unreliability' in place of `(Failure Rate X Time)'. Please explain this part.

[Lance]

Hi Rajesh,

Thank you for the additional information. Per your previous post, I first wanted to reiterate that there are no units assigned in the Xfmea software. Please remain consistent throughout your analysis. For example, if your measures are in terms of hours, convert all of your entries to hours and all of the outputs will be in terms of hours. Please remember that the Reliability characteristics of the Item/System should be included in this conversion of your inputs.

I would also like to remind you that the mean life can be shown instead of the lambda if you prefer. This is done by selecting User Setup from the File menu. Then click on the Other tab, and change the setting under Exponential Parameter as required before clicking on OK.

Getting to your second post, this is essentially the same calculation in Xfmea as in the Mil-Std 1629, Task 102. The Mode Ratio (Failure Mode Ratio) and the Probability of Loss are the same values in both. Additionally, when you are performing the Criticality Analysis, you are specifying a time value (t).

The only difference comes in the final term, which is failure rate in Task 102 and the Unreliability in Xfmea. The failure rate in the Mil-Std assumes a constant failure rate or exponential distribution. In Xfmea we allow you to specify a failure distribution, and do not limit you to the Exponential distribution. The reason for this is to allow you to specify reliabilities more like what you actually see in the field, and are not limited to a constant failure rate. Therefore, the Unreliability is essentially the failure rate for the given distribution at the given time.

[Rajesh Goyal]

Dear Lance,
In User Setup, Other Page GUI, there is an option to `Select Library’, since I have only evaluation package it is taking Defaults.lb3, but all I want to know is what it contains? In what way I will use them, or can you give us actual library which is being given with package so that we can evaluate the package with better understanding. No misuse assured.
- Rajesh

[Rajesh Goyal]

Dear Lance,
I am finding a problem in Xfmea-3 evaluation version. Problem is like this- When I am selecting profile MIL-STD-1629, and generating report, in ITEM SUMMARY reports following coloumn are not displaying any value:
Reliability,Fixed Reliability,Time-Dependent Distribution,Parameters,Data Source for Reliability. But If I select any other profile like Standard FMEA, it give all data which I have entered. Why it is so? can you find where I am doing mistake?
Rajesh

[Lance]

Hi Rajesh,

I hope that you also received our emailed replies to your questions. Regarding your first posting, the contents of the Defaults.lb3 library can be seen by selecting Profiles/Libraries Manager from the Tools menu. This is the same library that is shipped with the package when you purchase the software and includes as many of the existing standards as we could find documentation for (i.e. AIAG, J1739, Mil-Std 1629, etc…). If you are aware of any commonly used standards not in this list, please let us know.

I also wanted to point out that you are free to add your own standard or modify an existing profile/standard to include any additional terminology or scales used specifically by your organization. Then you can export the library file, and share it with the other Xfmea users in your organization.

Regarding your second posting, I wanted to thank you for bringing this to our attention. We have resolved the issue in our most recent Service Release. I apologize for any inconvenience that this may have caused you. Please download and install the Service Release from http://www.reliasoft.com/xfmea/updates.htm.

As always, please let me know if you have any other question or comments on the software. Thanks again for your feedback.

Regards,

Lance

[Rajesh Goyal]

I am facing one problem. When I am using Plot viewer and after drawing any Pie diagram and taking it to RS Draw for editing full diagram is not displayed: (a) Legends are not displayed with full text, some text is missing; (b) In pie charts if information is displayed and this is taken in RS Draw for editing normally, lower left information is lost. Even Zoom-in and out and refresh dose't solves the problem. I dont find any method by which I can place the Legends at my place of choice; Is there any solution for these. We are in the process of ordering the product, so please give solution/advice.

[Lance Larson]

Hi Rajesh,

Any information in the Plot should be transferred to the RS Draw tool. However, it seems as if you might have very long names for a Cause or Effect that might be truncated either on the plot or on the legend. If they are truncated in the Plot, then they will also be truncated when transferred to RSDraw. If you are seeing anything other than long names being truncated, then please email me with more specifics (screenshots are always welcome) at support@reliasoft.com

Thanks again for your post.

Have a great day!

Lance

[Rajesh Goyal]

Dear Lance,
This is not the case, I am geting it in the plot but they are truncated in the RS Draw. I can send the pictures also(In that case I will again need the evaluation copy and Licence), but you can go to the help file, in which also I found picture of Pie diagram which appears to be truncated. I am sure even if you take the example plot given on page 90 of xfmea_training.pdf, in RS Draw you will get the problem, because my descriptions are of approximately same length (Just put Legend and User Information also on. But, I am sure what I get in Plot Viewer, I don't get the same in RS Draw for edit.
Rajesh