Compare design alternatives before changing the accepted model
Use scratch previews, named configurations and read-only Review requests to compare changes with dimensions and evidence.
An alternative is easier to judge when its tradeoffs are explicit. “Make this stronger” could mean thicker walls, a rib, a different material or a changed load path. Those choices have different effects on mass, clearance, manufacturing and appearance. A good comparison begins with the question that matters.
Set a baseline
Open the accepted revision and record the relevant envelope, component placement and measured volume. Note which requirements must remain fixed. For an enclosure, the outside footprint, connector positions and mounting pattern may be non-negotiable even while wall thickness changes.
Volume is a geometric measurement. It becomes a mass estimate only with an appropriate density and material assignment. A larger volume is not, by itself, proof of greater stiffness or strength.
Ask Review for options
Press R to open Review. The panel shows the current revision’s measured context and audit status, and can submit read-only AI requests for accuracy checks, alternatives or realistic detail priorities. A useful request is: “Propose three ways to improve this mounting feature. Compare space, likely manufacturing implications and what evidence is still needed. Do not edit the model.”
Keep suggestions separate from completed engineering work. An AI proposal has not been built, checked or analyzed simply because it sounds plausible.
Preview numeric changes
For an existing parameter, ask for a scratch rebuild: “Preview wall thicknesses of 2.0, 2.5 and 3.0 mm. Keep the external envelope fixed. Compare measured volume and internal clearance without changing the saved revision.” Parameter previews perform a noncommitting build and return actual geometry facts.
Some dimensions may not be independently editable. A thicker wall could close a slot or violate a clearance. A failed preview is useful evidence that a dependency needs attention. Do not silently suppress that feature to obtain a successful comparison.
Save named configurations
When a numeric parameter set is worth retaining, ask for a named configuration. Configurations store parameter values, not a general promise that the design is valid under every combination. Applying one rebuilds against the current revision, and the resulting geometry still needs the relevant audit and requirement checks.
Use explicit sketch constraints where relationships must persist. A one-time arithmetic expression in a parameter experiment does not create a permanent dependency between dimensions.
Commit the chosen result and verify it
State which option to apply and which criteria remain unchanged. Ask for measured differences, a snapshot and any failed or unverified checks. Inspect front and side views when changes affect interfaces. Retain the baseline revision for comparison.
If the decision depends on stress, heat flow or motion, use an appropriate study with defined physical inputs. A parameter comparison cannot establish performance that was never computed. Kinematic playback cannot establish contact forces or stability.
Bring the comparison to a design review
Request a pinned-revision presentation with actual views, dimensions, audit context, a component register and factual notes. Label proposed improvements and unverified assumptions clearly. The deck should help the team decide what to test next.
Install once.
Design through conversation.
Download Intact v0.20.0 for Windows. Install privately in Codex with your own account; CAD runs on your computer.
Follow the installation procedure →