Some Thoughts on Wood Lathe Safety
After watching Sam’s video at Wyoming Wood Turner on lathe safety and watching Martin’s Turner’s Journey sharing of his recent accident, I decided to put out a copy of a report my undergraduate students produced on face shields. My interest developed after reading of a horrible accident in the AAW journal that a woodturner experienced even though she was wearing a face shield. The students came to my house, took a few lessons on wood turning and tried out some of the face shields I use. They then set out to study this further. What follows is their report. I edited out the equations because of the problems in translating the original PDF file into the blog. If anyone wants a copy of the complete PDF, please send me your email address.
I would welcome your comments on this study and will pass your comments on to the students. Of course there is a lot more to lathe safety besides the size and speed of the turning but maybe this will inspire turners to look for other means of protection or encourage manufacturers of face shields for wood turners to come up with new designs. Regardless of the type face shield we wear, there is no substitution for common sense. Stay out of the line of fire when possible. Turn under those cages that are sometimes provided with lathes in lieu of using them for tool racks. Use the best practice in mounting your blank to the lathe. Inspect your blank carefully for defects. Use your ears. If it doesn’t sound right, it may be wrong. Stop your lathe often and check for potential failures. Don’t wear rings or long sleeve shirts. I guess the list could go on and on.
MODELING IMPACT OF WOOD LATHE FAILURES ON FACE PROTECTION
BY CHAD OLNEY AND LAURA DETARDO
The process of shaping wood on a lathe, “turning”, involves applying significant forces by hand to material often of irregular surface profile and cross-section rotating at high velocity, and as such, precautions against accidents must be taken to avoid serious injury. In addition to wood shavings and debris common to all wood turning projects, imperfections throughout the wood can cause hand tools to shift unexpectedly and cause failure within the workpiece, presenting a risk for injury to the operator.
Faults within the wood including grooves, burrs, and rotted sections can cause a hand tool to break off significant segments from the material, and improper mounting of the workpiece on the machine can cause the workpiece to release from the chuck and collide with the operator. Even with the use of a guard or shield which can be fixed to the base of the lathe, there is a danger of large wood fragments being thrown from the lathe toward the lathe operator, and it is at the discretion of the operator what the degree of protection is required for a project.
Lathe operators typically wear eye or face protection while working on a piece to protect from reduced visibility and injury from airborne detritus. The aim of this report is to detail the results of a study of the ability of a commercially available face shield to absorb the impact energy from wood of various species and sizes at different spindle rotational velocities in the scenario of turning a bowl blank. This study was conducted in response to videos on the internet in which experienced woodworkers were wearing the recommended safety equipment but were still injured when struck by a sizeable fragment. Fragments from a workpiece can be propelled with sufficient energy to cause bruising, concussion, and even fatal damage to the head, even when the operator is wearing head protection which meets nationally-accepted safety standards.
It is posited that the majority of face shields available commercially are not adequately designed to absorb the energy of impact of a larger projectile. The ANSI (American National Standards Institute) standards which must be met by commercially available polycarbonate face shields state that the shield must be able to absorb 0.84 J of impact energy, while a “high-impact” shield must absorb 4.41 J . The ANSI standard is documented in ANSI/ISEA Z87.1-2010 . An experiment was devised to simulate the impact of a projectile from a wood lathe failure and determine the adequacy of the shield to absorb the impact, and further theoretical study was conducted to estimate the effectiveness of commercially-available face protection against wood lathe failures which could not be replicated in the previous experiment.
II. BACKGROUND RESEARCH
When researching the different types of face shields available for woodworking it was determined that there is two basic styles. The first involves just a headband that holds a plastic shield that covers the wearer’s face. The second, more common, shield combines a helmet and face shield, or provides some rigid structure or protection for the wearer’s forehead or around the sides of the plastic . This second style also comes with the option for a pressurized respirator to keep the wearer cool while working . Examples of both types of face shields are shown in Figure 1.
Fig. 1: Examples of both basic types of face shields used in woodworking  .
The American National Standards Institute (ANSI) provides a rating scale for each face shield. Under ANSI ratings, there are two protection levels; the lower level or basic impact, which must be achieved by all face shields sold commercially, tests the strength of the shield by dropping a 1 in diameter steel ball weighing 68 g from 50 inches onto the shield at the point which corresponds approximately to either of the wearer’s eyes . These hold an ANSI rating of Z87, which means the shield can withstand 0.87 Joules of energy during an impact.
The upper level or high impact rating utilizes a high velocity test where a ¼ inch-diameter steel ball bearing weighing 1.06 g is shot at 300 ft/s towards the face shield. An example of the recommended apparatus for consistently reproducing the high velocity test is shown in Figure 2. The Z87.1-2010 standard specifies that the 300 ft/s speed must be achieved no further than 25 cm from the point of impact. A high mass test is also used for rating for Z87+, in which a 500 g pointed projectile, the geometry of which is shown in Figure 3, is dropped from a height of 50 inches. If the face shield does not dent, crack, or displace from the frame, it earns the Z87+ rating .
Fig 2: The recommended apparatus for the high velocity impact test.
Fig. 3: The geometry of the high-mass test impact missile .
Corroborating the kinetic energy calculations for both the high mass and high velocity tests with other sources, it is known that shields that can withstand an impact of at least 4.41 Joules of energy hold an ANSI rating of Z87+, and must pass both the high velocity and high impact tests. However, it should be noted that the powered respirators may carry the Z87+ rating but can have a thinner piece of plastic than most face shields .
Further investigation into the background of the Z87.1 standard revealed that the parameters set by ANSI/ISEA to establish the safety ratings are largely arbitrary and that the organization itself is self-regulating, so examination is needed to ascertain whether a safety rating for “high impact” has any practical meaning for wood turners.
III. EXPERIMENTAL METHOD
The initial approach to this experiment was to build a pendulum that would generate the same angular momentum as a that of the rotating pieces of wood that could possibly break off while woodworking on a lathe. Three different wood types were chosen with four different diameter/length combinations to create cylinders, or dowels, which would form the impact head of the pendulum. The maximum angular momentums (tables seen below) were then calculated for those twelve combinations. The results of the calculations for each wood species/diameter/length/spindle speed combination are shown in Tables 1-3.
Upon further research, it was determined that using a pendulum would not sufficiently translate the energy generated from the pieces of wood into the face shield. The pendulum could not store enough energy to deliver the impact force at the desired magnitude, so an alternate experimental setup was devised. Rather than creating separate pendulums with different impact heads, an air cannon could be constructed using an air compressor, a butterfly valve, and a length of Schedule 40 steel pipe. Pressure from the compressed air is released when the valve is opened, pushing the projectile, a wooden dowel, down and out the length of the barrel.
Initially, a similar approach to the pendulum-based experiment was taken for dowel geometry relating to mass: each dowel would be the same length, and the radius of each dowel would be scaled from 0.5” to 2” in steps of 0.5”. The airflow would be restricted for the smaller sized dowels using an O-ring or rubber stopper. Initial mass and calculations were conducted with these parameters in mind . However, it was determined to be more efficient to have dowels of the same diameter to minimize pressure loss down the length of the barrel, so the dowels would be scaled by length to achieve the same mass as their diameter-based counterparts with a 2” diameter.
The revised experiment utilized a compressed air cannon launching wooden dowels at a target, measuring the exit velocity of the dowel from the cannon and calculating the force of impact. The equation to calculate the cannon exit velocity was derived by researchers from Wabash College assuming adiabatic expansion to obtain the following, where m is the mass, Po is the initial pressure at the valve, Vo is the volume of the cannon reservoir, A is the cross-sectional area of the barrel cavity, L is the length of the barrel, Patm is atmospheric pressure, gamma = 1.4 for air, f and is the friction factor :
Exit velocity =(4)
Because each dowel had the same diameter, only one cannon needed to be constructed to accommodate all of the different dowel lengths, so all of the parameters listed above are the same for the exit velocity calculation except for the mass of the dowel. Equation (4) was used to determine the size of a piece of wood needed to generate 4 Joules of energy at the instant of exit from the barrel. Only wooden dowels were considered for these calculations and ultimately tested.
The final set up for the air cannon used an air compressor with a 100 psig capacity connected to a 2 ft length of pipe with a 1 ¼” diameter. Several adapters were applied in series to allow the ¼” fitting of the air compressor hose to connect to the 1 ¼” pipe. To measure the velocity of the projectile, the cannon was set at a 45 degree angle from the ground on a ramp approximately three feet from the ground and secured as shown in the picture below. At this angle, an equation can be used to approximate the velocity of the projectile, where Vo is the exit velocity, d is the horizontal distance traveled by the dowel while airborne, and g is acceleration due to gravity.
Fig. 4: The unweighted 1 oz dowel (left) and the same dowel with the 2 oz weight added (right).
Fig. 5: The testing setup configuration for the air cannon during data collection.
The same dowel was fired from the cannon several times for data collection. The first tests utilized an oak dowel weighing one ounce, but due to the lightness of the dowel and the clearance between the dowel and the walls of the cannon, too much pressure was lost along the length of the cannon, and the exit velocity was considerably less than predicted as a result. The next series of tests used a 1 oz dowel with a 2 oz metal bolt inserted along the length of the dowel. The same 3 oz projectile was fired from the cannon five times, and each distance was recorded.
IV. RESULTS AND DISCUSSION
When testing the 1 oz. dowel, the average distance when shot at 45 degrees was 37 feet. Plugging those variables into the energy equation, we determined the 1 oz. dowel generated 1 Joule of energy. The 3 oz. dowel was then loaded into the air cannon and again shot at 45 degrees, the average distance was 47 feet and generated roughly 4 joules of energy.
To get a better understanding of what that 4 J of energy translates to, the air cannon was laid flat on the table and the 3 oz. dowel was shot at a sheet of cardboard that was held roughly 18 inches away. The picture below shows the results of that test. It should be noted that the hole in the center of the impact was due to the metal weight that was inserted into the center of the dowel to increase its mass, however the indented ring around that center hole was caused by the impact of the dowel.
Fig. 6: A photograph of the damage dealt to the cardboard sheet by the dowel. The dented, unpunctured section highlighted by the arrow is the impact of the dowel itself.
After reviewing the results of the experiment, it became known that there were scenarios within the scope of the study for which even faceshields rated for Z87+ impact would not adequately absorb the energy from an impact. Rather than replicate these impacts with further tests from the air cannon, a theoretical study was conducted to determine maximum safe conditions for turning operations in a variety of circumstances. Because the high velocity impact test described earlier is conducted with a horizontally-moving projectile, the impact energy is equated to the kinetic energy of the projectile. The impact energy is calculated at five rotational velocities for three species of wood and three blank diameters from which failures can eject. The volume of the projectiles being evaluated are rough approximations of the volumes of the two-inch diameter dowels used in the air cannon experiment which are detailed in Table 4.
The table cells colored blue in Tables 5-13 are those containing impact energy values for which a basic-level rated shield is sufficiently safe, those colored green are safe for Z87+ rated shields, and the red cells denote scenarios for which there is a risk for injury even when wearing a “high impact” face shield. The bottom row of each table contains the calculated RPM value for which it is safe to turn a workpiece on a lathe to be completely protected from the force of impact by a projectile of the specified volume. To provide some illustration of the volumetric dimensions, a cube of volume 2.5 in3 would have an edge length of approximately 1.35 inches, a 10 in3 cube has edge length of about 2.15 in (about 1.5x the size of the 2.5 in3 cube), a 20 in3 cube has edge length 2.71 in (2x the size of the 2.5 in3 cube), and a 40 in3 cube has edge length 3.42 in (2.5x the size of the 2.5 in3 cube).
The results of this study show that shields rated to only basic Z87 impact standards protect against wood lathe failures of small size and should be used only when turning at low spindle speeds. Shields rated to Z87+ will protect against impact from some larger failures at higher spindle speeds, but the range of safety coverage is still greater at low speeds for all woods. It is uncommon for wood turning operations occurring at spindle speeds like 2500 RPM to be much more than surface finishing, so the likelihood of encountering dangerous failures at such speeds is low as long as good woodworking discipline and general common sense are employed when working at the lathe. It can be seen from the tables that even a Z87+ shield will not protect the wearer from injury against especially large failure impacts.
With this information in hand, it is desirable to develop a face shield which will guarantee the safety of the wearer. Improvements can be made to current designs to increase the suitability of commercial face shields to withstand impact forces such as those delivered by the previously described wood lathe failures. For example, adding thickness to the polycarbonate shield itself will increase the rigidity of the shield, preventing it from deforming under the pressure of the impact. A riot helmet containing a face shield is subject to a separate safety standard set by the National Institute of Justice. In their standard NIJ-0104.02, the criteria for face shield impact testing is similar in process to that of ANSI/ISEA Z87.1-2010, but the impactor used to verify that the shield meets the standard is a 45-mm diameter cylinder with a weight of 1 kg being dropped from a height of 80 cm . This would suggest that the shield on a riot helmet must be able to absorb approximately 7.85 J of energy. Using this value in accordance with Tables 5-13 shows a significant increase in the range of the conditions deemed safe or minimally hazardous for lathe failure impacts. The thickness of riot helmet face shields is on average about 3/16”, while the polycarbonate on the face shields tested in the experiment ranged from 0.1-0.15 inches .
An alternative solution to increasing the ability of the face shield to protect the wearer is to incorporate existing methods of head and face protection used in other types of helmets, particularly those used in contact sports. Football and hockey helmets contain a lattice of thick wire bars which are used to distribute the force from impacts from objects such as hockey pucks, which are of approximately the same dimensions as some of the lathe failures analyzed in this report, which would be equally effective, if properly arranged, in protecting against impacts encountered in wood turning. These can be secured to the polycarbonate face shield to prevent the full contact area of the projectile from impacting the shield itself, and the remaining energy can be absorbed by the shield without significantly reducing the wearer’s field of vision. This would also reduce the risk of large wood projectiles denting or cracking the shield.
Sports helmets also include padding in the helmet itself so that when impact forces are encountered, the rigid plastic does not transfer the full force of the impact to the wearer’s head. Similar cushioning can be implemented into wood turning-appropriate face shields, particularly the forehead and potentially the chin, to mitigate the effects of a collision with a lathe failure projectile.
In this study, the ability of a face shield which was deemed able to protect its wearer from injury when impacted was tested against actual workpiece failure scenarios for wood turning operations on a lathe. Tests were conducted using an air cannon to determine the impact force of a wood projectile delivered to a shield and to set a basis for further theoretical analysis to better define the boundaries of safe operation for real-world wood lathe use. It was discovered that the application of the “Z87+” label to rated shields does not necessarily protect the wearer from highmass and/or high-velocity impacts, and data was tabulated to provide a clearer picture to wood turners to what degree they are protected by their headgear. Face shields rated for basic impact were found to be almost entirely inadequate for protection against all but the lowest-intensity evaluated conditions, and even those rated for higher impact did not cover the majority of scenarios. Further expansion of this study should include an analysis of how the energy calculations translate to degree of danger posed by the impacts to the health of the wearer at each condition, particularly which scenarios can prove excessively harmful or fatal, even when proper head protection is in use.
 A. Chen, “Are You Wearing the Right Faceshield, American Woodturner”, vol. 25:2, pp.
14, April 2013.
 A. Jackson, D. Day and S. Jennings, The Complete Manual of Woodworking, 13th ed., NY, Alfred A. Knopf Inc., 2012.
 A. Rao, Dynamics of Particles and Rigid Bodies: A Systematic Approach, 2nd ed., Cambridge U.K., Cambridge University Press, April 2011.
 H. Carpenter, “Safety for Woodturners: On the Edge of Disaster, American Woodturner”, vol. 27:4, 2013, pp.16.
 J. English, “Wood Dust Solutions, American Woodturner”, vol. 25:2, pp. 20, April 2010.
 R.G. Chandavale and T. Dutta, “Correction of Charpy Impact Values for Kinetic Energy of Test Specimen” in Pendulum Impact Machines: Procedures and Specimens for Verification, T.A.
Siewert and A.K. Schmider Eds., Philadelphia PA, American Society for Testing and Materials, 1995, pp. 221-231.
 Time-Life Books, The Art of Woodworking: Woodturning, Alexandria VA, St. Remy Press, 1994.
 W.H. Wagner and C.E. Kicklighter, Modern Woodworking: Tools, Materials, and Processes, 6th ed., South Holland IL, The Goodheart-Wilcox Company Inc., 1986.
 Z.J. Rohrbach, T.R. Buresh, M.J. Madsen, ” Modeling the exit velocity of a compressed air cannon,” Am. J. Phys., vol. 80, no. 1, pp. 24-26, January 2012.
 International Safety Equipment Association, “American National Standard for Occupational and Educational Eye and Face Protection Devices ANSI/ISEA Z87.1-2010,” International Safety Equipment Association, Arlington, 2010.
 “Adjustable Face Shield,” Harbor Freight Tools, [Online]. Available: http://www.harborfreight.com/adjustable-face-shield-46526.html.
 “Face-shield high temperature,” Direct Industry, [Online]. Available: http://www.directindustry.com/prod/jsp/face-shield-high-temperature-15890-591308.html.  U.S. Department of Justice National Institute of Justice, “NIJ Standard for Riot Helmets and
Face Shields,” October 1984. [Online]. Available: https://www.ncjrs.gov/pdffiles1/nij/097212.pdf.  “Riot Face Shield,” Gentex, [Online]. Available: http://www.gentexcorp.com/shopaviationhelmets/ground-accessories/eye-face-protection/faceshields/riot-face-shield.
Do not use any information in these posts without permission.