PYTHON · NUMPY · IN-BROWSER

Choosing a material and a process

For the sensor-mount part and a stated requirement (strength, cost, count), select a material+process pairing and justify it against at least two alternatives, using the twin's estimated cost, time, and feasibility.

01Challenge

Pick a material and a process that meet all four requirement lines for 5,000 brackets (count 5,000; survive 40 N; ≤ $3.50/part; first parts within 3 weeks). You haven't been taught how the families differ — try a pairing, estimate, read what the twin says, and iterate to a combination marked feasible and under cost at this volume.

02Model

There are really only three ways to make a solid thing: cut it out of a block (subtractive), pour liquid into a mold and let it form (molding/casting), or grow it layer by layer (additive). The trick isn't geometry first — it's count. Cutting and growing barely care about volume; each part costs about the same at one or a thousand. Molding starts brutal — you pay for a steel mold up front — then every part is almost free.

Somewhere there's a break-even; for 5,000 brackets you're well past it. Count picks the neighborhood; then geometry and material pick the house. So you don't pick a process — you pick a pairing, and justify it by beating the alternatives on the requirement.

Process cost crosses over with volume.

03Guided practice

Pass A (worked): Aluminium 6061 + CNC @ 5000 → ~$11.40/part, feasible but far over $3.50; CNC cost barely drops with volume → wrong family.
Pass B (hint): for 5,000 you want the FORMED family — a moldable thermoplastic + injection molding. Check strength (40 N), temp, and unit cost < $3.50.
Pass C (independent): lock a pairing AND fill the justification matrix — name ≥2 alternatives and the requirement axis each loses on.

04Feedback

PASS WHEN Chosen pairing is feasible and meets all four requirements (strength, cost, count, lead), and ≥2 alternatives are each beaten on a losing axis that matches the twin's estimate, on seed 1102.

Aluminium + CNC is feasible but unit cost ~$11.40 > $3.50 — CNC cost barely falls with volume; at 5,000 parts move to the FORMED family (thermoplastic + injection molding).
PLA + 3D printing is cheap on screen but fails strength: PLA can't meet 40 N + outdoor temp. Keep molding, switch to ABS or Nylon-6.
Your pairing passes but the justification names only 1 alternative (need ≥2) and no losing axis — for each alternative, name the requirement it fails.
You wrote Aluminium+CNC 'loses on strength' — it's actually stronger; it loses on cost ($11.40). Match the losing axis to the numbers.

05Retrieve & space

From Lesson 1.1: your process is injection molding — which tagged defect does a mold care about most? → no draft.
From Design course: in the part header, identify a functional dimension (fixed) vs a shape dimension (free) — pre-loads 1.3's function-preservation rule.
Spaced: drag the injection-molding break-even marker to roughly where molding overtakes CNC on the count axis.

06Mastery & project

In sim, select a pairing the twin marks feasible and meeting all four requirements, with a justification beating ≥2 alternatives whose losing axes match the twin's estimates, on seed 1102.

Commits the part to a process (ABS + injection molding), which decides which DFM rules 1.3 must apply, which process M2 generates toolpaths for, and the cost basis the M5 capstone optimizes.

← What "manufacturable" means Designing for the process (DFM) →

= REQ['load_N'] and m['temp_ok']\n return dict(unit=round(unit_cost,2), days=round(days,1),\n feasible=feasible, strength_ok=strength_ok, notes=notes)\n\nfor pair in [('Al6061','cnc'), ('PLA','fdm'), ('ABS','molding')]:\n print(pair, estimate(*pair))","label":"2 — Deterministic estimator"},{"code":"# EDIT: your chosen pairing, and >= 2 alternatives each with the axis it LOSES on.\nchoice = ('ABS', 'molding')\njustify = [ # (material, process, losing_axis in {'cost','strength','count','lead'})\n ('Al6061', 'cnc', 'cost'),\n ('PLA', 'fdm', 'strength'),\n]\ne = estimate(*choice)\nprint('chosen', choice, '->', e)","label":"3 — Your choice + justification"},{"code":"def meets(material, process):\n e = estimate(material, process)\n weeks = e['days']/5.0\n return (e['feasible'] and e['strength_ok'] and\n e['unit'] <= REQ['cost_target'] and weeks <= REQ['lead_weeks']), e, weeks\n\nok, e, weeks = meets(*choice)\nproblems = []\nif not ok:\n if not e['strength_ok']: problems.append(f'{choice[0]} fails strength/temp for the 40 N + outdoor requirement.')\n if e['unit'] > REQ['cost_target']: problems.append(f'unit ${e[\"unit\"]} > ${REQ[\"cost_target\"]} target at 5,000 count.')\n if weeks > REQ['lead_weeks']: problems.append(f'lead {weeks:.1f} wk > {REQ[\"lead_weeks\"]} wk.')\n# validate justification: >=2 alternatives, each truly losing on its named axis\naxis_val = {'cost': lambda x: x['unit'], 'strength': lambda x: 0 if x['strength_ok'] else 1}\njok = len(justify) >= 2\nfor mat, proc, axis in justify:\n a = estimate(mat, proc)\n if axis == 'cost' and not (a['unit'] > e['unit']): jok = False; problems.append(f'{mat}+{proc} does not actually lose on cost.')\n if axis == 'strength' and a['strength_ok']: jok = False; problems.append(f'{mat}+{proc} does not actually lose on strength.')\n\nif ok and jok and not problems:\n print(f'PASS - {choice[0]}+{choice[1]} meets all four lines (unit ${e[\"unit\"]}, strength OK, '\n f'lead {weeks:.1f} wk). You beat 2 alternatives on their losing axes. A defended pairing.')\nelse:\n print('FAIL - ' + ' '.join(problems[:2] if problems else ['Need >= 2 alternatives, each with a real losing axis.']))","label":"4 — Autograder (seed 1102)"}],"intro":"The twin's cost/time/feasibility model is a deterministic function of (material, process, count, geometry); here it is in numpy. Edit your chosen pairing and run.","key":"manufacturing/choosing-material-and-process","kind":"python","title":"Choosing a material and a process"}">

PYTHON · NUMPY · IN-BROWSER

Choosing a material and a process

The twin's cost/time/feasibility model is a deterministic function of (material, process, count, geometry); here it is in numpy. Edit your chosen pairing and run.