Thursday, August 18, 2011

Biofouling–A life of its own.

Is your system ready to take on a new life? Many technology solutions require clean water for their process. In the semiconductor industry, ultra-high purity water is a must with extreme deionization of water. Residential water heaters even have problems with scaling, and many of the high-efficiency tankless systems cannot be installed in areas with hard water.

Scaling issues with calcium and magnesium are commonly on the mind of a system developer maybe because it is something we experience everyday. On the other hand, biofouling is not always considered until it is too late. Much like scaling, once biofouling starts it may be impossible to treat, control or recover from the growth.

In system testing, it is common to set specifications for the end user and test for the environment that is readily available. Water quality around the world is not equal.

Crazy_Red_BugA Tale of Two Microbes

A problem was experienced when multiple systems tested in the Boston Massachusetts area over several years had no indication of problems with water except for hardness, which was treated with standard filtration and ion exchange materials. A major market was a well developed European country. The problem started with system overheating that required a tweak to the cooling pump curves every few months to keep the system operational. When a shutdown occurred and the system

Tuesday, August 16, 2011

Process Temperature Sensitivity

Temperature sensitivity of processes should be considered early in the design before specifying equipment. Frequently, designers specify a single thermal measurement device for ease of use, replacement, and inventory. Not a bad idea if all temperaturTemperature_Displayes are near the same range or the process is not sensitive. K-Type thermocouples are very popular due to their wide range of temperatures (-200˚C to 1250˚C), but when measuring high temps of 1000˚C and water temperatures around 50˚C what is the error associated with them?

Many designers do not realize that an error even exists and even process interfaces and data files will be set up with one or two decimal places. The standard error of a K-Type thermocouple is a minimum of 2.2˚C and at temperatures nearing 1000˚C the error nears 7˚C. So with even a 4.4 degree spread on error, reading temperatures in tenths of a degree is useless, requires more data-storage  space and adds irrelevant

Wednesday, July 20, 2011

Design for Usability–Computer Design Failure

According to research performed by Deliotte, 2012 will be the year to exceed 1 billion computers produced annually. With 500 million units shipped in 2009 you would think that the computer industry would have things fairly well worked out for design, but do they? Consumer usability is always an issue and frequently overlooked. Take a few new cars for test drives and you are bound to say “I really don’t like where that thing is placed.”

computer faceHere is a PC that is just a few years old.

Look at the power switch.

It seems that every time a disc is placed in the lower drive the computer shuts down and needs to be rebooted. Sometimes when the headset cords are plugged in, a wild finger will cause a reboot. It also happens when installing or removing a thumb drive in the USB port. Very frustrating to have your system shut down from attempting normal use.

Yes, the computer

Friday, June 24, 2011

Design for X (DfX), Excellence or ‘X’ as a Variable?

“When you say DfX or Design for X, is the ‘X’ intended to be for "’excellence’ or the use of X as a variable as used in an algebraic equation?’” This question was asked of me when explaining what I do. “Well, it is dependent on the person you ask or where you look for the definition. If you use it as a variable it should cover the term ‘excellence' as well as any other term you want to use. Let’s go with ‘X’ as a variable.” was my response.

Regardless of the term you use for ‘X’ there are certain realities that are inescapable. If a system cannot  be

Monday, April 11, 2011

Failures and Unintended Component Applications

Extending the limits and the specification boundaries of components is common in innovative environments, as often a standard component does not exist yet for a new technology.  In this case, a pump designed to deliver low temperature liquids was used in an application due to its specified flow curve, operating pressures, and cost. The vendor specification did not

Monday, April 4, 2011

Compressor Filter Failure

This small compressor has a filter located on the upper port. The function of this compressor was critical to system operation and life time. Several systems using these compressors ran well for long periods, but ultimately a unit failed.  The post mortem analysis discovered a small piece of material in the outlet Compressor_Filtervalve, and further investigation showed degradation of the filter material.
What went wrong? The filter, in this application, was being used was an exhaust filter designed for outbound flow., while the original configuration of the compressor was to be used as a vacuum pump. The person specifying the filter was unaware that the default setting set by the vendor was as a vacuum pump, nor that one filter could be configured in two ways. could be configured in more than one way.  
Prior to this failure, another problem plagued the system: an orifice manifold located near a hot entry point was being clogged with small brown chucks of material that were thought to be reversing through the system when the compressor was not running. The clogged orifice caused severe operating problems and resulted in the degradation and reduced the life expectancy of a downstream component nearly one quarter of the total system cost. The system was redesigned with expensive check valves to eliminate the reverse flow situation. After the compressor filter was found to be an issue, testing of old orifice was found to contain melted fibers from the filter and not material from back-flow.  That same filter caused these problems as well.
Concurrent engineering practices that involve persons who understand the consequences of even the simplest failures can reduce errors like this. The proper use of tools such as an FMEA may have identified this as a potential failure, but required a person knowledgeable enough to identify it. The FMEA in this case identified the compressor as a component and did not take the filter itself into consideration. It is critical to involve the right people early in the process to reduce failures and hidden lifecycle costs Blogger Labels: Case Study,FMEA,Compressor,Filter,Failure

Wednesday, March 30, 2011

A Breakdown in Communication

A communications cable was manufactured by a custom cable and harness supply house, and due to use, the housing had been removed along with the heat shrink so as to expose the blue and white wires.  What was visible is that the white wire on this cable is exposed; what you could not see was a single strand from the white wire hovering dangerously close to the blue wire. As the system vibration frequency changed within operation modes, the close proximity of the blue and white wires would cause an intermittent fault.  The result of the fault, considered by itself, was not too severe: the system shut down.  The resulting system-wide damage was costly, however, due to the thermal stress of multiple shutdowns.  Additionally customer perception declined quickly.COM_Cable_1

The failure is in the design, rather than in the manufacturer. The cable specified a wire gauge too large for the connector, causing the close proximity of the blue and white wires. An off-the-shelf bulkhead connector was specified due to cost and availability, while a custom design with smaller gauge wires would have cost more initially, it would have saved service time, product life, and customer perception.  How could the team have chosen the correct connector up front? Manufacturing drawings, tied together with an electrical design review, could have brought this problem to light prior to deployment.  While the hardware cost for the correct connectors was $100, compared to $5.00 for the bulkhead connectors, the overall cost of using the bulkhead connectors exceeded $8,000 due to travel expenses, service engineer cost and the cost of the software development tool to detect the problem.  The frustration level of the customer was elevated enough to call off the installation of another system Blogger Labels: case study,Failures,Breakdown,Communication,system