BlockSim Example: CM Triggered by Subsystem Down

Purposes
The purpose of this example is to illustrate how to model the case when CM is triggered by failure of subsystem. We present two kinds of methods: method without subdiagram and method with subdiagram. The limitation of the method without subdaigram is that, it cannot model subsystem with more than two devices in parallel or more complex structure. The method with subdiagram can break this limitation.

Method 1: Without subdiagram
Let’s consider a system with three devices A, B and C are in parallel. A and B form a subsystem AB. Only both A and B fail (subsystem AB fails), then start to repair A and B. Device C is parallel with subsystem AB. The repair of C depends on its failure. When C fails, then start to repair C.

BlockSim Solution
The BlockSim modeling of this system is shown in Figure below.



Reliability of all blocks follow Weibull distribution with Beta =1.5 and Eta=500 hours.

Block A and Block B are belonging to maintenance group MG_AB.

CM of Block A and B are set to "When found failed during an inspection". The CM duration is 120 hour. Upon repair, they are as good as new.

There is an inspection for both block A and B. The inspection is performed based on events in maintenance group MG_AB. If "Block fails" or "Start of corrective maintenance" event occurs for any of the blocks in the maintenance group MG_AB, then the inspection is performed. The inspection duration is set to "Immediate Repair" and upon inspection; the device is restored to same as it was when it failed (as bad as old).

CM of Block C is performed upon failure of Block C and the duration is 20 hours. It is as good as new upon repair.

Block Up/Down plot
The system event log is shown in Figure belwo and is as follows:


 * 1) At 271, Block C fails and get repair at 291.
 * 2) At 331, Block B fails. Since the subsystem AB is not down (Both A and B fails, then subsystem AB fails), Block B is waiting for repair opportunity after its failure. (Failure of Block B cannot trigger inspection of Block B, but this failure would trigger the inspection of Block A. However, Block A doesn't have failure at this time, thus nothing happens).
 * 3) At 381, Block A fails and subsystem AB fails. Subsystem fails trigger the inspection and then inspection triggers CM. (failure of Block A triggers an inspection on Block B, and Block B has failure at this point. Thus CM of Block B is performed. Performing CM on Block B triggers the inspection of Block A and Block A also has failure at this time, thus CM of Block A is performed too.)
 * 4) At 501, both Block A and B are restored.
 * 5) At 791, Block B fails and is waiting for repair opportunity.
 * 6) At 943, Block C fails and get restored at 963.
 * 7) At 1358, Block A fails and subsystem AB fails, triggering CM for Block A and B.
 * 8) At 1478, both Block A and B are restored.



Method 2: With subdiagram
Let’s a system consists of four devices A, B, C and D. Devices A, B and C are in parallel structure and form a subsystem ABC. Only all A, B and C fail (subsystem ABC fails), then start to repair devices A, B and C. Device D is parallel with subsystem ABC. The repair of D depends on its failure. When D fails, then start to repair D. Method 1 cannot model this system. Here we use subdiagram to model the subsystem.

BlockSim Solution
The BlockSim model of this system is shown in the Figure 5 below. By using subdiagram, we can break the limitation in method 1.

Reliability of all blocks follow Weibull distribution with Beta = 1.5 and Eta = 100 hours. Subdiagram block ABC belongs to maintenance group MG_ABC. CM of Block A, B and C are set to "When found failed during an inspection". The CM duration is 20 hour. Upon repair, they are as good as new. There is an inspection for Block A, B and C. The inspection is performed based on events in maintenance group MG_ABC. If "Block fails" event occurs for any of the blocks in the maintenance group MG_ABC, then the inspection is performed. The inspection duration is 30 hours. Upon inspection, the device is restored to as bad as old. CM of Block D is performed upon failure of Block C and the duration is 40 hours. It is as good as new upon repair.



Block Up/Down plot
The system event log is shown in Figure belwo and is as follows:


 * 1) At 13, Block C fails. Since subsystem ABC doesn't fail, Block C waits for repair opportunity.
 * 2) At 111, Block B fails. Since subsystem ABC doesn't fail, Block B waits for repair opportunity.
 * 3) At 176, Block A fails and brings down the subsystem ABC. Failure of subsystem ABC triggers the inspection of Block A, B and C.
 * 4) At 197, Block D fails and get repair at 237.
 * 5) At 206, Block A, B and C finish inspection and trigger CMs. CM is performed on Block A, B and C.
 * 6) At 226, Block A, B and C are restored.