A high quality parabolic drill might speed it up.
Perhaps over 50%.
You might use a short drill to some depth, then another to finish the hole.
As was said breakage and parts cost might be important, or not.
S:
Ask iscar to help and bring drills to try.
Drill sample holes in one sacrificial part of same material to prove the process, with iscar present.
As was said, don´t try for top speed.
If 35 secs/part is achievable, 80 secs might be much more stable, and usually less tool wear.
The HSS drill probably heats in the hole in 3 mins.
This heats it, jamming it more.
Exponential heat/friction occurs.
S:
Could you dam the part via say plastic walls, flooding it under coolant.
This might allow pecks to keep the drill cool.
S:
Or perhaps full retracts on pecks vs short rapid pecks.
This might allow evacuating chips better.
A VMC can retract the drill really fast, and you wont lose time, but the hole gets evacuated and the drill tip gets cooled.
S:
Is the hole anywhere near straight uniform and cylindrical ?
Gage pins or shop built gages can easily show you if it is somewhat straight.
Or drill rod of 3 mm, and another of say 3.05 mm, -- 0.01 mm incremental for a no-go.
3 mm is very flexible and can only take little push, and HSS is relatively limp spaghetti.
I am pretty surprised and very impressed if the hole is anywhere near straight, uniform and cylindrical.
Please let me know where do buy, what drills, if the holes are straight (ish).
The stock answer is solid carbide drills.
And perhaps bore the start by about 3-4 D or about 12 mm in this case after stub predrilling about 2.7 mm D.
With a say 8 sec toolchange, you are maybe 8 drill + 8 + 8 bore + 8 + final drill == 30 secs, 62 secs total.
The bore would support the drill and tend to reduce breakage.
The start would be very straight and uniform.
A bored start hole, or perhaps just solid carbide stub drill, might reduce breakage to near zero.
My SWAG is around 90 secs, bit less, is somewhat easy to achieve with good reliability re: tool breakage.
My SWAG is that somewhere under 60 secs is doable if you really tune and test the process and some breakage is acceptable.
How will you detect broken drills with short cycle times ?
Could you contact probe, perhaps independently, via the part and FH or stop or estop as needed via a relay.
The O. post was about production, and these types of things tend to come up.
Imho, imhe.
Just throwing things out there, that might perhaps help.
Weather you are doing pallets of 400 parts, or one at a time, and they need accurate fixturing, or not, and you have a Heller or top japanese machine, or a Johnford etc. may affect the best minimax.
2000 / week or 4000 may be a major critical path issue.
Since you did not mention it I expect hole accuracy is not an issue at this time.
This is unusual and surprising with HSS at 17D deep drilling into alloy steel.
If the hole is a clearance hole or assy hole of no particular accuracy ...
.. and the part is not particularly expensive, so wastage is ok, ..
fast drilling with solid carbide, maybe deep pecks to clear chips under coolant, might be the right route.
If you are really doing production volumes, you can easily add through-tool coolant to the machine for a few thousand.
This should reduce times by maybe 30-40 secs from 90 ish to 50 ish.
So 2-3000 / week could become 5500 / week.
But the hole alone may not be the bottleneck at 5k, or maybe it is.
At 3 mins/hole, thats 200 parts max per day/11-12 hours, or 1000 ish/wk.
At 90 secs, potentially 2000 /wk.
At 35 secs, 5000 theoretical.