Scope the mechanism before printing

The cheapest change to a printed mechanism is the one made before the first print. Once parts exist, there is a strong pull to make them work: sand this, drill that, add a shim. Each fix makes the next build harder to repeat. A short written scope, an hour’s work, prevents most of that by deciding in advance what the mechanism must do, what conditions it must survive, and which jobs printing should and should not be asked to do.

This chapter scopes the six-station indexing table that runs through the book. The method applies to any printed mechanism: a demonstration gearbox, a camera slider, a test rig, a classroom model or a functional prototype.

Start with the job, not the shape

A scope begins with what the mechanism must achieve, stated in numbers wherever possible. “A rotary table that indexes” is a wish. “Six stations, 60° per step, about ten steps a minute, a pause of a few seconds at each station, positioned within half a degree” is a specification that can be designed to and tested against.

This is an illustration. The indexing table’s functional requirements are:

Figure 1Functional requirements for the indexing table

6 stations60° per indexdial of 140 mm diameter

About 10 per minuteIndex rateadjustable down to about 4

4 sDwell at each stationat full speed; longer when slowed

±0.5°Station positionabout ±0.48 mm at the 55 mm station radius

200 hRunning lifeas a teaching model

The position tolerance converts to a distance at the radius where it matters. Stations sit on a 55 mm radius, so ±0.5° there is 55 × 0.5 × π ÷ 180 ≈ ±0.48 mm. That is a number a builder can check with a dial indicator or even a ruler and a marked pointer.

The running life matters more than it first seems. At ten indexes a minute for 200 hours, the table makes 10 × 60 × 200 = 120,000 indexes. Chapter 5 shows that the small sun gear in the planetary stage meets its three planets so often that each of its teeth is loaded about 1.8 million times over the same life. Printed gears can manage that at light load, but only if the design respects it.

Estimate the loads early

Even a teaching model has loads, and the scope should estimate them before any part is drawn. The estimates need not be precise. Their job is to show which parts are lightly loaded and can be designed for shape, and which carry enough force or torque to need a calculation.

This is an illustration. The gearmotor chosen for the table has a data sheet that lists a rated torque of 0.15 N·m at 50 rpm and a stall torque of 0.45 N·m. Through the 5:1 planetary stage, with an assumed efficiency of 90%, the rated torque at the Geneva driver becomes 0.15 × 5 × 0.9 = 0.675 N·m. If the dial jams and the motor stalls, the driver sees 0.45 × 5 × 0.9 ≈ 2.03 N·m.

The Geneva drive pin sits 30 mm from the driver axis, as Chapter 6 calculates. The force on the pin is the torque divided by that radius: about 0.675 ÷ 0.030 = 22.5 N at rated torque and about 2.03 ÷ 0.030 ≈ 67.5 N at stall. Those are modest forces, but they act on a small pin and a narrow slot, which is why the pin is steel and the slot is checked later.

The stall case matters more than the rated case for most printed parts. A mechanism that jams, because a student’s finger, a pencil or a loose screw gets into it, will see stall torque until someone switches it off. The scope therefore records two load cases: rated running and stall. Chapter 5 checks the gears against both, and Chapter 8 adds a current limit so that the stall case is brief.

Decide who will build it

A mechanism built once by its designer can rely on the designer’s knowledge: which way a part goes, how hard to press a bearing, which hole needs reaming. A mechanism built by others, from files and a guide, cannot. The scope should say which it is, because the answer changes how much the design itself must prevent mistakes.

This is an illustration. The table will be built by teachers and senior students, often without the designer present. That decision drives several later choices: parts that can only fit one way, numbers and arrows printed into parts, an alignment tool for the Geneva timing, checks after each subassembly, and a build guide tested by someone who has never seen the table. Chapter 9 brings these together. Writing the decision into the scope means each chapter can design for it from the start rather than adding warnings at the end.

Write down the conditions

Printed polymers are sensitive to temperature, moisture, sunlight and sustained load in ways that metals are not. The scope should state the conditions the mechanism will meet.

• Temperature. Where will it be used and stored? A classroom is one thing; a car boot on a summer day in Australia is another. PLA softens at around 60 °C, its glass transition temperature, and a parked car can get hotter than that. PETG’s glass transition is around 80 °C. These are typical values; check the data sheet for the actual filament.
• Sustained load. Polymers creep: under a constant load, they slowly deform. A screw clamping two PLA parts loses some of its clamp over weeks. A printed spring loaded continuously relaxes.
• Handling. Will students take it apart and reassemble it? Repeated assembly wears threads cut into plastic and fatigues snap fits.
• Sunlight and cleaning. Ultraviolet light degrades some plastics over time; cleaning solvents attack others.

This is an illustration. The table will be used indoors, stored in a cupboard, carried between rooms and reassembled by students each term. It will not be left in a car, but the scope notes the risk, because one hot afternoon could warp a PLA frame.

Decide what printing should do

Material-extrusion printing is very good at some jobs and poor at others. A mechanism that plays to its strengths works better and lasts longer.

Figure 2What desktop printing does well and badly in a mechanism

	Good at	Poor at
Shape	Complex frames, ribs, curved guards, part consolidation	Very flat, very straight or very round surfaces
Accuracy	Repeatable features once calibrated	Precise holes and shafts straight off the printer
Strength	Stiff enough shapes at low load	Interlayer strength, sustained load, impact in the weak direction
Wear	Occasional sliding at light load	Continuous sliding, bearing surfaces, threads used often
Iteration	Cheap, fast design changes	Matching a bought part's tolerances without measurement

The practical rule is print the shapes, buy the precision. Shafts, bearings, pins that carry load or wear, springs and threads that will be used repeatedly are usually better bought. Frames, brackets, gears at modest loads, guards and housings are good candidates for printing.

Make the print-or-buy list

This is an illustration. Applying the rule to the indexing table produces a short list.

Figure 3Print-or-buy decisions for the indexing table

	Decision	Reason
Frame, dial, covers	Print, PLA	Shape-driven, lightly loaded, indoors
Planetary gears and Geneva wheel	Print, PETG	Shape-driven; PETG for toughness and heat margin
Shafts	Buy, 8 mm steel	Straight, round and stiff; printed shafts would bend and wear
Bearings	Buy, 608 ball bearings	Low friction, long life, standard size
Geneva drive pin	Buy, 6 mm steel dowel	Sliding contact in the slot, the highest wear point
Threads	Buy, M3 heat-set inserts	Students reassemble the table; printed threads would strip

The Geneva drive pin is a good example of a small decision with a large effect. The pin slides in and out of a slot six times per dial revolution, under load, for 120,000 indexes. A printed pin would wear flat; a steel dowel pressed into the printed crank costs very little and removes the problem.

Choose materials for each part

Within printed parts, material choice follows the conditions and the job. PLA is easy to print accurately, stiff and widely available, but has the lowest heat resistance of the common filaments and can be brittle. PETG is tougher, has more heat margin and is less prone to brittle fracture, but strings more and is slightly harder to print to tight dimensions. Other materials, such as ABS, ASA, nylon or fibre-filled blends, offer more heat resistance, toughness or wear resistance in exchange for harder printing, fumes or cost.

This is an illustration. The table uses PLA for the frame, dial and covers, where stiffness and accuracy matter and loads are light, and PETG for the gears and Geneva wheel, where toughness and a little more heat margin are worth the extra printing effort.

A note on emissions: printing releases fine particles and volatile compounds, with the amount depending on material and temperature. Print in a ventilated room, follow the filament supplier’s safety data sheet, and prefer an enclosure with filtration where printers run for long periods near people.

Standardise the hardware

Every different screw, insert or bearing in a design is another item to buy, store, sort and get wrong at assembly. Standardising hardware early simplifies everything downstream.

This is an illustration. The first layout of the table used seven different screws: M2.5 and M3 in several lengths, plus two self-tapping sizes. A review reduced them to three: M3 × 8, M3 × 12 and M3 × 20 socket-head screws, all into M3 heat-set inserts. The builder then needs one hex key, one insert size and three bags of screws, each a clearly different length.

Figure 4Hardware variety before and after standardising

Screw types, first layout7 types

Screw types, standardised3 types

Insert types, standardised1 types

Bearing types, standardised1 types

The three lengths were chosen to be easy to tell apart at a glance: 8 mm, 12 mm and 20 mm differ by at least 4 mm, so a builder holding two screws side by side cannot confuse them. Lengths that differ by only 2 mm, such as 10 and 12, are a common source of wrong-screw errors, and a 12 mm screw driven where a 10 mm one belongs can bottom in its insert or punch through a thin wall. Where the design needed an in-between length, the boss or clamp thickness was adjusted to suit one of the three standard screws instead of adding a fourth. Each screw position in the CAD assembly is labelled with its length, and that label carries through to the build guide, so the choice is never left to the builder’s judgement.

The same logic applies to bearings and shafts. All four bearings in the table are 608s, a common size that is cheap and easy to find, and all shafts are 8 mm to match their bore.

Set a budget and a test plan

The scope should finish with two things: what the mechanism may cost and how success will be judged.

This is an illustration. The budget is about A$150 including GST for one table’s materials and bought parts, and about 30 hours of print time. Chapter 9 itemises it. The test plan follows the requirements directly: count indexes per minute at full speed, time the dwell, measure station position at all six stations over several revolutions, run the table for an extended soak and check for wear, and have someone other than the designer build it from the guide.

Figure 5Scope sign-off for the indexing table

• Avoid: Functional requirementssix stations, rate, dwell, position, life
• Avoid: Conditionsindoors, reassembled termly, not left in cars
• Avoid: Print-or-buy listshafts, bearings, pin and inserts bought
• Avoid: MaterialsPLA frame, PETG gears and Geneva wheel
• Avoid: Hardwarethree screw lengths, one insert, one bearing
• Do: ? Gear life at 1.8 million sun-tooth cyclesto be checked in Chapter 5
• Ask: Station position ±0.5°depends on Geneva and dial fits, Chapters 3 and 6

Keep the scope where the files are

A scope written once and filed away is soon forgotten. Keep it in the same folder or repository as the CAD files, the print profiles and the build guide, and give it a revision number. When a requirement changes, such as a faster index rate or a new material, update the scope first and then follow the change through the design. Each chapter of this book adds evidence to the same folder: coupon measurements, fit decisions, gear checks, wiring diagrams and test logs. By the end, the folder explains not just what the mechanism is but why each choice was made, which is what the next designer will need.

Recognise a scope failure

This is an illustration. An earlier teaching model in the same workshop, a printed winch, had no written scope. Its designer printed the drum, gears and frame in PLA and used printed pins as axles. It worked well in the workshop. After a week in a sunny window display, the drum had sagged on its axle and the gear teeth had deformed where they had been held under load by the hanging weight.

The cause was two unstated conditions: sustained load and heat. A scope would have asked both questions, and the answers, steel axles and a display away from the window, cost almost nothing. The lesson is that the scope’s job is to ask the questions that the workshop environment does not, because a mechanism that works on the bench has passed only the bench test.

The edited Library chapter on notes on prototyping discusses what a prototype is meant to learn, which is a useful frame for deciding how much scope a printed mechanism needs.

Exercise: scope a faster table

A colleague wants a version that indexes 30 times a minute for a production-line demonstration, running 8 hours a day for 40 days. How many indexes is that, how does it compare with the teaching table’s life, and which items in the scope would you revisit first?

Worked answer. At 30 indexes a minute for 8 hours a day over 40 days, the table makes 30 × 60 × 8 × 40 = 576,000 indexes, nearly five times the teaching table’s 120,000.

Revisit, in order:

1. Gear life. If the planetary stage runs three times faster to reach 30 indexes a minute, every gear tooth sees three times the cycles per hour as well as nearly five times the total cycles. The PETG gears need a fresh check in Chapter 5, and bought or moulded gears may be more sensible.
2. The Geneva pin and slot. Three times the speed means nine times the inertia torque, because acceleration scales with the square of speed, and more impact at engagement. Chapter 6 covers this.
3. Heat. A faster motor working harder runs warmer inside a printed enclosure, and PLA near a warm motor may soften. Consider PETG or ventilation for the motor mount.

The rest of the scope, such as the hardware list and the print-or-buy decisions, probably holds.