In my June 2015 blog, I discussed a personal injury case where the jury decided that the lack of interlocks on the covers of a swamp cooler did not render it defective, even though it was technologically feasible to do so and their existence would have prevented the amputation. Conversely, in another interesting personal injury case involving a computer numerically controlled (CNC) milling machine, a large settlement was made by the manufacturer with the injured operator because the machine did not have adequate interlocks on its enclosure doors.
The milling machine in question was being used to manufacture heads for a large racecar engine. The operator was attempting to drill long holes through a ten-inch thick block of aluminum. A deep hole drill bit with through-the-bit coolant holes was used to drill the holes. Coolant was forced through holes in the bit near the tip in order to keep the tip lubricated and cool, and also to flush metal chips out of the hole being drilled.
The operator programmed the milling machine to run at 8000 RPM, the speed he understood to be recommended by the tool manufacturer for the material he was drilling. After setting up the part and the tool, he opened the enclosure doors several inches apart to enable him to see that the tool was turning in the correct direction and that coolant was, in fact, coming out of the holes in the drill point near the tip of the bit.
The CNC machine was initially designed with an interlock switch connected to enclosure doors that would not allow the machine spindle to operate faster than 750 RPM with the enclosure doors open. However, when the machine was initially installed, the manufacturer’s installer told the operator that the machine’s controller was programmed with a control system override. If one line in the program was changed by the installer, the operator would be able to override the door interlock, open the doors and observe the tool rotation and coolant flow, no matter the rotation speed of the spindle. The installer said he knew operators wanted to do this. At the request of the operator, the installer made the line change in the program, allowing the operator to override the door interlock and operate the machine spindle to its programmed speed, as high as 10,000 RPM, with the doors open.
The operator did not realize that the end of the deep hole drill bit needed to be supported at the tip to prevent it from deflecting excessively due to centrifugal forces. If a long bit is rotating off center just a little bit, centrifugal force tries to bend the tip outward away from the centerline. The more the bit deflects and the faster the bit turns, the higher the force which causes more deflection and higher force. If the bit turns fast enough, it will break and the broken tip can become a missile that can cause serious injury or death. A risk assessment performed by the manufacturer actually indicated that the machine had an unacceptable risk of tool ejection beyond the recommended 750 RPM maximum speed.
On the day of the incident, the operator pressed the “START” button and the milling machine spindle went from 0 to 8000 RPM. The bit was not perfectly aligned, as it never is, and the centrifugal force on the unsupported drill bit caused it to deflect and break. The broken tip was ejected from the mill through the gap between the enclosure doors. The bit hit the operator in the head right between his eyes, fracturing his safety glasses frame. He suffered a traumatic brain injury and damage to his optic nerves. In addition to having anger outbursts from the traumatic brain injury, he was declared legally blind and in need of full time care the rest of his life. This operator’s life will never be the same.
It was my engineering opinion that this machine should never have had the ability for the door interlock to be overridden. A risk assessment for this machine indicated that the severity of harm could be catastrophic and the probability of occurrence of harm was very high, which resulted in an unacceptable risk of harm.
One thing I had to do was reconcile the issues associated with this milling machine opinion with the previous opinion I posted about the swamp cooler in the previous blog. I did that by comparing the two machines in the table below. (Fig 1) I concluded that the reason for opening the guard for the CNC machine was to check the rotation of the tool and whether or not the coolant was coming out of the tip. For the swamp cooler, the guard was opened for cleaning. The guard on the CNC machine needed to be opened daily or more often to check the tool rotation and coolant flow. The swamp cooler had to be cleaned only every two to three months. The CNC machine needed to be running. The swamp cooler did not need to be running and, in fact, could have been unplugged. With all of the above noted, I concluded the risk of serious injury or death was high with the CNC machine, but low for the swamp cooler.
Risk assessment is an important step in the machine design process.
- Just because a guard can be interlocked does not mean it should be.
- It is not the manufacturer’s responsibility to make a machine that is risk free, but instead reasonably safe.
- The need for an interlocked guard depends on whether the risk is tolerable.
In summary, every machine has risk. There is no such thing as zero risk. Manufacturers are expected to estimate the risks. There are well-accepted processes to estimate risks. Reasonably safe means that any residual risk is tolerable. Manufacturers have a responsibility to use well accepted processes to estimate the risks of the machines they produce, and to ensure their machines are reasonably safe.
Jeffery H. Warren, PhD, PE, CSP, is the chief engineer and CEO at Warren specializing in mechanical, machine design and safety. His deep expertise in machine design and safety analysis makes him a frequent presenter, trainer and expert witness. In addition to investigating more than 2000 claims involving property damage and injuries related to machinery and equipment since 1987, Jeff has an undergraduate degree in Mechanical Engineering from the University of North Carolina as well as a Master of Science and a Doctorate in Mechanical Engineering from Virginia Polytechnic Institute and State University — both with machine design emphasis.