Download No-Cost Raptor Software

Section 1-
RAPTOR software is a modeling software program for:
            Reliability (tells you how many failures to expect),
            Availability (tells you how much uptime to expect), and
            Maintability (tells you about maintenance down time and times between maintenance demands)
A RAM model tells you numbers for how your system operates:

1.     Yes, I know you’re an expert about things such as pumps, compressors, furnaces, heat exchangers, and vessels!

You’ve also got information about their ages to failure and times for repairs.  You can also organize the data into a statistical format for use in a RAM model to forecast how well things will perform in the future.

It’s important to make your data speak for future events in building a model for new process and equipment. We need the facts.

Other web pages at tell you how to reduce these facts into the few details you need for the RAM model.

2.     Yes, I know you have great engineers who fit the system together with expertise about how the blocks are conveniently arranged for a production process. 

These same “fathers” or “mothers” of the process often will look you in the eye and deny the deteriorating effects of time and loads that cause entropy changes resulting in failures of “their systems—think of their children”.  Our provincial outlook makes it difficult to admit the facts about how the creations we conceived will fail. 

All systems fail—the key is when, where, and how often.  All systems require maintenance.  We can’t repeal the natural laws of entropy changes.  This means we must account for the failures and the effects of the failures as the model reflects real-life conditions.

3.     Yes, I know you have experts in the process and how the system should perform on a steady-state basis.  Unfortunately we design processes and equipment for steady-state conditions.  However, we operate the systems in a dynamic mode.  What we design and how we operate them often incurs problems which we need to predict.

Systems deteriorate and systems fail so use the facts from your records to gain details for your models.

4.     What you rarely know, on a factual basis, is how the installed system will really function for number of failures, availability, and maintenance demands.  The bottom line of these numbers must be reflected into time and money.

Monte Carlo RAM models allow you to compute the cost of unreliability (COUR) for a business perspective.  They are helpful in searching for cost effective alternatives for life cycle cost considerations.

Few engineers can support the ideas in items 1-4 with engineering calculations.  They often resort to opinions--not facts.  For business results we need calculated facts before we build so we can convert the facts into $ statements.  We must make business decisions from the technical details. 

Why model? 

·       We need more operating facts summarized into numbers and less opinions devoid of numbers!

·       We need operational details before we commit to bricks and mortar in a plant site. 

·       We need the facts to match the economics for the plant or process. 

·       We need these calculated facts from the model on one side of one sheet of paper.

·       We need a practical representation of our operating plans into a computer simulation model so we can watch it run under simulated conditions. 

·       RAPTOR helps work out the details using a Monte Carlo model (allows randomness in how the equipment operates before it fails and it also allows randomness in how long it takes to repair equipment) so we can visualize what’s happening to our overall system and components in the system. 

·       Data for RAPTOR models should come from your own maintenance data systems.  Weibull databases are available to represent life/death of equipment.  Other data sources are available in the reading list on this website with data sources marked by a dagger symbol, .

RAPTOR is useful for engineers for converting technical details into money for life cycle cost considerations.  It does not require statisticians to build the model or run the model.  It lets you perform “What If?” exercises to arrive at the best tradeoffs for the system.  Best of all, the system produces numbers by simulation using your facts—not opinions.  You can learn what is going wrong and where you can find some advantageous over your competitor by building the plant and operating it as a computer model before you install the equipment.  Rearrangement of plants and processes by cutting torch is an expensive process compared to building a simple RAM model.  Remember the techie details are not the primary important issue in most models—the money is what’s important; and time/money is what the business must consider—of course we use the technical details to get to the money!

You can download some technical papers (as PDF files) showing RAPTOR models and life cycle costs from this site—just follow the hyperlinks below.  Three papers listed below show how RAPTOR has been used for making life cycle costs decisions using the factual details from the field of reliability engineering and processes to arrive at the lowest long term cost of ownership (this is concept important to the stockholders view and requires that engineers must think like MBA’s as they act like engineers):

1.     How To Justify Equipment Improvement Using Life Cycle Costs and Reliability Principles

2.     Why You Need Practical Reliability Details To Define Life Cycle Costs For Your Products and Competitors Produces

3.     How To Justify Machinery Improvements Using Reliability Engineering Principles

Other life cycle cost papers are available for download as PDF files or you can go to the short list at the bottom of the LCC training page.

Section 2-
You can down load the old RAPTOR reliability block diagram software, at no cost.  These old software programs were produced by the US Air Force.  Two good reasons to use the handy RAPTOR software for reliability block diagram modeling using Monte Carlo techniques: 
  1) It’s free, and
  2) A novice can use the software after 10 minutes of play-time which makes the software very productive. 
On this site you can download two older Air Force copies at no-charge.
            RAPTOR Version 2.99 ßdownload only for the help file. (Read the RAPTOR Acronyms at ).  Or you can simply download the help file by clicking here.
            RAPTOR Version 4.0S ßSee Section 3 below (for better software). 

Version 4.0S is the last version without a copyright produced by the Air Force and some report problems running this version of Windows7. (Note: you can get later and better versions of no-cost demonstration software, remember demo software is not the same as the full version of software which you must purchase.  The demo version and the full version are now produced by Booz Allen Hamilton as described below in Section 3).  Please note the lack of downward file compatibility by version—upward file compatibility is OK.

Section 3-
ARINC moved the former US Air Force software into the commercial area with their greatly improved version 7.0.  In 2013, Booz Allen Hamilton Engineering Services purchased RAPTOR from ARINC.    

You can download a demonstration version of RAPTOR version 7.0 by simply registering with the Booz Allen site.  A comparison table of the features of different version of 7.0 is available.

You can download a tutorial from .  You can also download a product sheet for RAPTOR version 7 from .

Section 4-
A couple of hints on how to get started with RAPTOR--
            First thing to build a model, in versions below 5.0, after you open RAPTOR is to put a START node on the left hand side, then put a STOP node on the right hand side of the grid. [This start/stop node sequence is not mandatory for version 7 as it will accept the insertion at any time] 
Do this by clicking on EDIT, click on ADD, click on NODE and select the proper node (by the way, I always add extra CONNECT nodes because it seems the number of blocks grow like Topsey and you should freely add intermediate CONNECT nodes as it simplifies your life.  Exit the nodes selection by right mouse click.

            Second thing is to add blocks for each element in your model. 
Do this by click on EDIT, click on ADD, click on BLOCKS and put them into a logical/functional diagram.  If you have “no clue” as to the life distribution, always choose EXPONENTIAL and select the Mean, i.e., MTTF (leave the location as zero) and select LOGNORMAL  and select Mean, i.e., MTTR with the default standard deviation (actually it’s a shape factor for the lognormal as it has no dimensions because it’s a lognormal value) of 2 units.  When you’ve added the correct number of blocks, right mouse click to exit the block addition.  By the way, all units for MTTF and MTTR must be consistent—let’s assume you use hours!!  You can get some ideas about repair times from the example in the July ’01 Problem Of The Month.

            Third thing is to add links. 
Do this by click on EDIT, click on ADD, click on LINKS, and then click with drag at the center of the start node and while holding down your left mouse key, drag the link to the next node/block.  Repeat until the model is linked together.  Exit the link mode by right mouse click.

            Fourth thing is to click on FILE (now is a good time to save your file) and click on SIMULATE. 
If you can’t get to the simulate button, try a right mouse click!  A dialog box will ask for your STOP time i.e., mission time—let’s assume it’s 8760 hours.  The dialog box will also ask for the NUMBER of simulations—choose 100 or more because not every block in most diagrams will have a failure in a single year and you should work on the basis of “safety in numbers”.

            Fifth thing is to follow your nose for the balance of the details you will need. 
As a memory jogger, when you select Weibull distributions for failure modes:
            the shape factor = b = slope of the Weibull line and
            the scale factor = h = characteristic age to failure, and
            the location = t0 = correction of the time origin. (Set this value to zero unless----see below.)
The correction of the time origin can be to the left (i.e., failures begin before you put the equipment into service) or to the right (i.e., a “failure free interval” prevents failures before a given time).  The usual value for location = t0 = correction of the time origin should be zero.  To get a correct value for t0 you will need:
            1) a sample size of at lest 21 failures to document a t0 value, plus
            2) a physical explanation of why this condition is true and
            3) other three parameter details outlined in section 3 of The New Weibull Handbook
So, the bottom line is to set location = t0 = 0 unless you have overwhelming evidence that supports your hypothesis!

            Sixth thing is convert the results from the model into time and money.
Everyone understands time and money (we usually have too little of both!).  Don’t give your audience a root canal by endlessly speaking of the technical details—in the end, it’s all about time, money, and alternatives—this is what you must talk about! to Barringer & Associates, Inc. homepage

Last revised 02/13/2014