Friday, 2:14 PM
Friday afternoon. Our phone rang, and the caller ID was an aerospace customer we'd worked with for six years.
We have a problem.
I'm a production manager at a precision CNC metal machining shop. My main job is triaging rush orders, deciding what's possible, what's not, and what it'll really cost. I've handled 200+ rush jobs in nine years, including same-day turnarounds for aerospace clients. But this one was tighter than most.
They needed 14 Inconel flow bodies by Monday at 6 AM for a qualification test. Normal lead time: four weeks. The previous vendor had quoted a low number on Tuesday, then added a 'complexity surcharge' on Friday, $9,500 on top of a $12,000 base. The customer finally said no, which is why they called us at 2:14 PM on a Friday. I've learned to ask what's NOT included before what's the price. That call was a textbook example of why.
The geometry was ugly. Deep internal channels, a 30-degree unsupported overhang, and a sealing face that needed to hold pressure. We could machine most of it, but not all of it in 36 hours.
Why Velo3D Was on the Table
We installed a Velo3D Sapphire about 18 months before that call. I was skeptical at first. Everything I'd read said Velo3D was for complex production parts, not for emergency orders. In practice, that Friday, it was the only option that gave us a shot.
Part of the reason we took Velo3D seriously was the velo3d spacex supplier connection. SpaceX doesn't publish long lists of approved vendors, but Velo3D has cited SpaceX and Anduril as customers for flight-critical components (Source: Velo3D customer announcements, velo3d.com). If a supplier can handle launch hardware, they can probably handle a qualification test part. That's not a guarantee, but it's a meaningful filter.
Our CNC lead pulled up a CNC tool holder diagram while we debated. It's a good reminder of how much of precision machining depends on the tool being held at exactly the right angle, with exactly the right extension length. The part we needed had an overhang that would have required a custom-angle head and a fixture we didn't have. Velo3D removed that constraint. Not all constraints, but enough.
The Worst Saturday
We decided to build on the Sapphire. I went back and forth for about an hour before that decision. On paper, CNC made sense: we already had the Inconel bar stock, and our 5-axis machine was idle. But my gut said Velo3D because of the internal channels. A CNC approach would need at least three setups and a custom EDM electrode. The additive route would need one build and one finish pass. On paper, CNC felt safer. My gut said additive was the only way to make Monday.
After I hit 'confirm' on the build, I kept second-guessing. What if the layer adhesion failed? What if the recoater hit the part? I approved the build at 4:30 PM, then went home and didn't sleep.
During the wait, a client texted a question that seemed random: 'When do you see results from CO2 laser?' They were used to laser cutting, where you load a sheet and see parts in minutes. I get it. With Velo3D, you don't see anything for hours. The build completes, then you stress-relieve, wire EDM off the plate, and finish-machine critical surfaces. You feel like you're waiting forever.
By 11:07 PM Saturday, the first part came off the build plate. I remember the timestamp because I texted it to our customer. I want to say the full build took under 14 hours, but don't quote me on that. It wasn't done, but it existed. That's the moment I started to believe the deadline was possible. I'm not 100% sure why that particular build ran clean, either. Maybe the powder, maybe the alignment, maybe luck.
Monday, 5:52 AM
We delivered at 5:52 AM, eight minutes early. The customer later told us the parts passed the qualification test. The quote we gave on Friday was the final price. No complexity surcharge, no unexpected support removal fee, no last-minute add-ons. That transparency is why they called us in the first place.
I used to think Velo3D and CNC were competitors. It took me 18 months and about 40 builds to understand that they're not. Velo3D made the CNC faster by giving us a near-net shape we could finish-machine in one setup. The CNC made the Velo3D acceptable to a quality engineer who needed a machined sealing face. In my opinion, they're teammates.
Would I recommend Velo3D for every rush order? No. On a simple bracket, a CNC machine is still faster and cheaper. But for a part that can't be fixtured, or has internal channels, or needs unsupported overhangs, Velo3D can be the difference between a Monday delivery and a Monday conference call with the customer's legal team. At least, that's been my experience with unusual geometries.
If you take one thing from this, ask the pricing question before the emergency happens. And if you're comparing process speeds, 'when do you see results from CO2 laser' is the wrong question, at least for additive. Ask what the part geometry requires. The clock starts when the file is verified, not when the build starts.