3D printing and how to use it to minimize BCT and per-unit-cost.

The word 3D printing in context of mass manufacturing of mechanical or electronic parts tends to get thrown around alot lately. All the flash and bang about 3D printing being hyped as the catch all savior to industry 5.0 seems more fantasy than fact. While 3D printing is a very useful modern technique for manufactuing parts, it still fails to beat the champion parameter of everything manufactuing  - Bottom Cycle Time (BTC).  

To explain this topic a bit better, imagine a competition between a Production line with just FMD 3D printers. And another line with the good old Plastic Injection Moulding. The former line represents the modern frontier of manufacturing technology - 3D printing. And the latter line represents the ol' reliable. We shall be comparing the advantages and disadvantages of each gen of Manufactuing tech used and try to find that sweet spot between the two. Our test part will be a 450mm quadcopter frame. A commonly used UAV frame in FPV Kamikazed applications, made popular by the recent Russia-Ukraine Conflict. 

If you're in the market for the fastest Fused Deposition Molding (FDM) , perhaps you have heard about the Creality K1 Max. With the blazing fast speed of 600mm/s and a wig snatching acceleration of 20,000 mm/s^2, the K1 max is NOT playing games. It is one of the , if not the , Fastest FDM printer you can buy. It can even brag about beating most of the industrial grade FDM printers in the market in terms of print speed.

ofc bigger and more robust ones do exist that are meant for industrial use, but this analysis only looks at the print speed parameter and not other parameters that may be better suited to industrial setting such as robust control mechanism and control system.

For convenience's sake, we shall assume the following for the:

• Size: 450 mm diagonal (~400×400 mm footprint)

• Type: Hollow, skeletal, mostly perimeter and sparse infill

• Wall thickness: 2.5 mm

• Infill: 15%

• Material: PLA 

• Print volume: Approx. 600–800 cm³ of material used

• Estimated weight: 700–900 grams

• Layer height: 0.2 mm

• Nozzle size: 0.4 mm

• Shells/perimeters: 2–3

• Speed: 250 mm/s average

• Acceleration: 20,000 mm/s² (high, allows quick ramps)

• Volume: 700–800 cm³

• Layer time: 15–20 seconds

• Total layers: 300–400 (at 0.2 mm layer height)

 Layer height = 0.2 mm  

Z height of frame = 30 mm  

Number of layers = 30 mm / 0.2 mm = 150 layers  

Extrusion width = 0.48 mm  

Extrusion area = 0.48 mm × 0.2 mm = 0.096 mm²  

Extrusion path length per layer = 5,300 mm  

Volume per layer = 5,300 mm × 0.096 mm² = 508.8 mm³  

Total volume = 508.8 mm³ × 150 layers ≈ 76,320 mm³ = 76.3 cm³  

(Adjusted estimate for sparse quadcopter frame)

Average print speed = 250 mm/s  

Time per layer = 5,300 mm / 250 mm/s = 21.2 s  

+ 5 seconds (travel + retraction + cooling)  

Effective time per layer = 26 seconds  

Total time = 150 layers × 26 s = 3,900 seconds = 65 minutes = 1.1 hours  

Add margin for slow layers and cooling = 45–60 minutes  

Woah, almost an hour on the high end just to make one part. That is a BCT of 1 hour. A horrible parameter for a part meant for mass production. There are also other disadvantages to consider, such as :

• Any kind of filament feeding issue such as blockages, broken filament, burnt heat element, etc can contribute tremendously to the BCT which is already very high. BCT should ideally be as low as possible for a part as simple as a quadcopter frame.
  
  
• Immense amount of waste. As high as 10% filament material wasted for making support for part.

On the contrary 3D printing bring the following advantages to the conversation:

• More control over improvements in production process flow.
• Faster part design changes and implementation in production.
• Low cost for producing new model batches as same printers can print many different types of models. High ROI on initial machine cost.

That was all I have to say about the future. Let's head over to the ol' reliable- Injection Molding.

Injection Molding is of 2 types - 

1. Plastic Injection Molding (PIM)
2. Metal Injection Molding (MIM)

Here are a few assumptions I have made for the PIM process, to make our lives easier:

• Mold fabrication: 2–4 weeks (CNC or EDM)

• Cycle time per part: 30–90 seconds

• Post-processing: Often none

• Volume required for payoff: ≥1,000 units

Once the mold is ready:

Per part: 1–1.5 minutes
For 100 parts: 2–3 hours total production time

But the tooling time + cost is the catch:

• Initial delay: 2–4 weeks

• Expensive upfront, cheap per part

There you go folks: 

It's a BCT of 60min vs 1.5 minutes.  

Disadvantages:

• Less flexibility in updating part design
• ROI on machines is based on the size of the mold use. A bigger machine for a small part can be better otherwise the only alternative is using a greater number of smaller injection machines by slicing sections of parts into multiple processes, which is inherently the more expensive solution.

So, what now?

Is 3D not for REAL manufactuing jobs? Is PIM the PIMP of Industry 5.0? Does inhaling the fumes from PIM give you superpowers? (It does not, this is a joke. Don't inhale plastic fumes :))

The simple answer is NO!!!!!!!!!!!!!!!!!!!!!!!!!!

There in-fact does exist a solution using both FDM and Injection molding together. 

• FDM can be used for rapid prototyping and evaluating physical design of WIP part.
• The printer part costs more, but since production run is only meant for validation it won't be made in mass quantities. 
• Once design is finalized, it will be sent to a MIM specialist who will use the part to create a negative and add necessary elements to create a metallic mold.
• This mold will then immediately be put to use on the PIM Production line, minimizing time spent in making molds without MIM and 3D printing.
• In the end, we enjoy the precession and flexibility of 3D printing and use of MIM in creating molds for the final PIM process.

This process ensures flexible, fast, reliable, and cost-effective manufacturing of parts with the minimum BCT as possible.