When a 3mm Drill Became a Week-Long Problem

The Job That Looked Simple on Paper

I had a job drilling 3mm holes through 200mm of titanium 6Al-4V. The L/D ratio was about 66:1, which is well within the range of what a gun drilling machine can handle. I had drilled similar holes in steel without any trouble. The customer needed four holes in a part used for an actuator housing. The tolerance was ±0.05mm on diameter and the straightness requirement was 0.1mm over the full length. Nothing about this job looked unusual when I quoted it.

The material was titanium, which I had drilled before. Titanium is sensitive to heat because it retains about 80% of the cutting heat in the tool, unlike steel which sheds most of the heat into the chips. That means coolant delivery is everything on a titanium job. I calculated the required coolant pressure using the formula for minimum flow through a gun drill coolant hole. The 3mm drill had a 0.8mm coolant hole, which gave me about 5 liters per minute at 80 bar. That should have been adequate for the material removal rate I was targeting.

The first few holes went fine. I drilled through the first 50mm without any issue. The chips were coming out as fine segments, the spindle load was steady at about 60% of maximum, and the coolant pressure was stable. By the time I reached 100mm depth, everything still looked good. At 150mm depth, the drill broke.

Five Drills in One Day

I removed the broken tool, installed a new one, and the second drill broke at the same depth of 150mm. I went through five drills before I figured out the problem. Here is the pattern I saw:

Each drill cost $80. The five broken drills cost $400. The lost production time cost more than both combined. The machine was down for a day while I ran tests. The job that should have taken four hours took an entire week from start to finish.

I spent the first day swapping drills. I spent the second day checking the machine alignment. I checked the spindle runout with a dial indicator. It was 0.002mm, well within spec. I checked the collet condition. No visible wear. I checked the coolant pump pressure at the machine inlet. The gauge read 80 bar. Everything looked fine.

The Worn Bushing I Could Not See

On the third day, I sat down and thought about the failure pattern. The break kept happening at the same depth range. That suggested a fatigue mechanism, not an overload. A fatigue failure in a small-diameter drill is almost always caused by vibration. The question was where the vibration was coming from.

I pulled the guide bushing out of the spindle and looked at it under a 10x magnifier. The inside diameter was supposed to be 3.005mm, giving a 0.005mm clearance with the 3mm drill. Under the magnifier, I could see a wear mark on one side of the bushing bore. The wear was about 0.01mm deep and covered about 60 degrees of the circumference. I could not see this wear with my naked eye. It looked like a perfectly good bushing until I put it under magnification.

The issue was a worn guide bushing. The bushing had 0.01mm of wear that I could not see without a magnifier. The wear was causing the drill to wobble at entry. The wobble was small at the surface, maybe 0.005mm of runout at the bushing exit. But as the drill went deeper, that small wobble amplified into a whipping motion. At about 150mm depth, the whipping exceeded the fatigue limit of the carbide drill, and the tool snapped.

The wear happened because the previous job on that machine used the same bushing for a different material that was more abrasive. The operator did not record the bushing wear in the machine log. I checked the logbook. There was no mention of bushing inspection for that tool change.

A new bushing cost $60. The five broken drills cost $400. The lost production time cost more than both combined. I replaced the bushing, ran a test hole, and the drill went through 200mm without any issue. The spindle load stayed at 55% for the entire depth. The surface finish was excellent.

The Inspection Procedure I Use Now

I now check guide bushings with a magnifier before every small-diameter job. I use a 20x illuminated magnifier that sits on the tool bench next to the machine. I follow a specific inspection protocol:

First, I clean the bushing bore with a solvent to remove any cutting oil residue. Second, I examine the bore under magnification, rotating the bushing to check all 360 degrees. I look for wear marks, scoring, or edge chipping. Third, I check the bushing ID with a pin gauge. If the pin gauge goes through with more than 0.002mm of play, I replace the bushing. Fourth, I record the inspection result in the machine logbook, including the measured clearance.

For drills smaller than 5mm diameter, I have added a pre-production test. I drill a test hole in a scrap block of the same material at the same depth. If the test hole comes out within tolerance and the drill shows no unusual wear patterns, I start the production run. This adds fifteen minutes to the setup time. It has prevented two more tool breakage incidents in the past six months.

The Physics of Wobble Amplification in Deep Hole Drilling

The failure mechanism I encountered is worth understanding in more detail because it applies to every small-diameter deep hole drilling job. When a drill enters the workpiece through a worn guide bushing, the entry angle is not perfectly aligned with the spindle axis. The angular error at entry, even if it is only 0.01 degrees, creates a lateral force on the drill body. That lateral force causes the drill to deflect. As the drill goes deeper, the deflection increases because the unsupported length of the drill grows.

The relationship between entry error and deflection at depth follows an amplification curve. For a 3mm drill with a 66:1 L/D ratio, every micron of runout at the bushing exit translates to roughly 10 microns of runout at the drill tip. The amplification factor is approximately linear with the L/D ratio. At 66:1, a 0.005mm entry error produces about 0.33mm of tip deflection. That is enough to exceed the fatigue limit of a 3mm carbide drill after a few seconds of cyclic loading.

I calculated the numbers after the fact and put them into a table for reference:

This table now sits on the wall next to our guide bushing inspection station. Any operator running a drill smaller than 10mm diameter with an L/D ratio over 30:1 knows to be especially thorough with bushing inspection.

Why Operators Skip Bushing Inspections

After this incident, I asked the other operators on my shift why they did not inspect guide bushings more frequently. The answers were revealing. Some said they did not know the bushing could wear in a way that was invisible to the naked eye. Others said they assumed the bushing was fine because the previous job had run without issues. A few admitted they were in a hurry to start production and did not want to spend ten minutes pulling the bushing for inspection.

I realized that the system was set up to encourage skipping the inspection. The bushing was inside the machine, behind a cover plate that required a hex key to remove. The inspection tool was a handheld magnifier stored in a drawer on the other side of the shop. The logbook was at the supervisor’s desk. Every barrier between the operator and the inspection made it less likely to happen.

I fixed all three barriers. I moved the magnifier to a bracket mounted on the machine enclosure. I attached a hex key to the bracket with a cable so it could not walk away. I moved the logbook to a shelf next to the machine control panel. The time required for a bushing inspection dropped from ten minutes to three minutes. The inspection rate went from about 30% of jobs to over 90% within a month.

Key Takeaways

• A 0.01mm wear mark on a guide bushing, invisible to the naked eye, caused five drills to break at the same depth
• The fatigue failure from bushing-induced wobble took 150mm of drilling depth to manifest. The failure mode was invisible until the tool snapped
• A $60 replacement bushing solved a problem that cost over $400 in broken tools and a week of lost production
• Small-diameter deep hole drilling is unforgiving of guide bushing wear because the L/D ratio amplifies any entry error
• I inspect every guide bushing under 20x magnification before starting a small-diameter job now, regardless of whether the bushing looks good to the naked eye
• A pre-production test hole in scrap material catches problems before they scrap production parts
• The machine logbook is only useful if operators actually record their inspections. I enforce this with a sign-off sheet now
• The barriers to inspection matter as much as the inspection itself. Moving tools closer to the machine increased inspection rates from 30% to 90%