Catalonia PAU Technical Drawing (Dibuix Tècnic II) common mistakes
Catalonia PAU Technical Drawing (Dibuix Tècnic II) · Understand the traps and how to avoid them.
When a tangency problem gives a point of tangency on a circle and asks for the tangent circle of given radius, the candidate draws the centre anywhere on the perpendicular to the tangent line at that point and then adjusts the radius by eye, instead of locating the centre at the exact distance r from the given point along that perpendicular.
The candidate remembers 'the centre lies on the perpendicular at the point of tangency' but forgets that the distance along that perpendicular is fixed by the required radius, so the construction is left under-determined and finished by trial drawing.
What to do instead
Write the locus list before drawing: for a tangent circle of radius r touching a given circle at a given point P, the centre is on line OP (O centre of the given circle) at distance r from P, on the side determined by internal or external tangency. Mark that point with a construction line and verify the second condition (distance to the other given element equals r) with a compass check before inking.
In an ellipse/circle affinity (afinitat) transformation, the candidate maps points using the wrong direction of affinity or the wrong axis correspondence, producing a transformed figure that is mirrored or shifted relative to the axis.
Affinity is taught as 'parallels to a direction, distances proportional to the axis', and candidates mix up which of the two conjugate directions belongs to the original versus the transformed figure, especially when the axis is oblique.
What to do instead
Fix a convention and state it on the sheet: label the axis, mark one pair of corresponding points A and A', and derive the affinity ratio k = (distance of A' to axis)/(distance of A to axis) once. Then map every further point by drawing through it a line parallel to AA' and applying the same ratio with the compass, measuring perpendicular to the axis only.
When constructing the intersection of a plane with a polyhedron in dihedral projection, the candidate joins visible vertices directly on one view without transferring the intersection points to the other view, so the two projections of the section are inconsistent.
The section is drawn as a 'shape' on the view where it looks convenient, and the candidate treats the second view as a copy rather than as the projection of the same 3D polygon.
What to do instead
Work point by point: for each edge, find where it crosses the plane (using a change of plane or an auxiliary plane containing the edge), mark the point in both views immediately, then join only points that lie on the same face. Finish by checking that every section vertex lies on the corresponding edge in both projections.
In a change of projection plane (canvi de pla) or rotation, the candidate transfers heights from the wrong reference line, so the new projection is displaced vertically even though its shape is correct.
The rule 'departures are preserved in a vertical-plane change, heights in a horizontal-plane change' is applied from memory without keeping a single, clearly labelled reference line (LT or new LT) as the datum for all measurements.
What to do instead
Before starting, draw and label the original ground line and the new ground line, and mark the datum direction with an arrow. Transfer each height with the compass from the original ground line only, and mark each transferred point with a small tick so a mis-transferred point is visible immediately.
When drawing the development (desenvolupament) of a cylinder or cone cut by an oblique plane, the candidate uses the true lengths of the cut edge measured on the front view instead of the true lengths obtained from the generatrices, giving a development whose boundary curve is wrong.
The cut ellipse appears as a straight line in one view, and the candidate reads segment lengths from that line, forgetting that those segments are foreshortened except where the generatrices are parallel to a projection plane.
What to do instead
Divide the base circle into equal parts, number the generatrices, and take each generatrix's true length from the view where it appears in true length (or by rotation). Plot each point at (arc length along the base, true generatrix length) on the development, then fair the curve through the plotted points with a French curve.
In a section view (tall/cut), the candidate hatches two adjacent parts with the same line direction and spacing, or omits hatching on a cut web, so the parts merge visually and the drawing is marked as non-compliant.
Hatching is treated as decoration applied at the end, and the candidate does not plan which parts are cut and how many distinct parts appear in the section.
What to do instead
Before hatching, list the cut parts and assign each a distinct hatch pattern (direction or spacing) per the standard; hatch only material actually cut by the cutting plane, leave webs and ribs unhatched when cut longitudinally, and keep hatching clear of lettering and dimension figures.
Dimensions are placed on the drawing without regard to functional reference: the candidate dimensions from a random edge instead of from the datum face, repeats the same dimension in two views, or omits the overall dimension, so the set is over- or under-dimensioned.
The candidate dimensions each feature as it is drawn (drawing order, not functional order), and never reviews the complete set for redundancy or gaps.
What to do instead
Identify the functional datum first and dimension from it; dimension each feature once, in the view where it appears clearest; then audit the set: every feature dimensioned exactly once, no closed chains, overall length/width/height present, and no dimension that can be derived from others.
In an isometric or cavaliera axonometric, the candidate measures lengths along the axonometric axes at their true value for all three axes, or draws ellipses for circles by eye instead of constructing them from the isometric square.
The scale reduction along the axes (and the 1:1 simplification allowed in isometric drawing practice) is confused with true length, and circles in oblique planes are drawn freehand.
What to do instead
State the scale convention at the start (true isometric with 0.816 reduction, or isometric drawing at 1:1 along the axes). For each circle, draw the enclosing square in the axonometric plane, mark the four tangent midpoints, and construct the ellipse through them, checking the major axis is perpendicular to the plane's normal direction.
When producing orthographic views from a pictorial drawing, the candidate places the top view below the front view or draws a hidden line as a solid line, and the views do not align in projection.
First-angle and third-angle projection are mixed, and the candidate draws what is visible in the pictorial rather than projecting each edge and deciding visibility from the front view.
What to do instead
Write the projection system on the sheet before drawing (first angle: top view below front view; third angle: above). Project every vertex with construction lines from the front view to fix alignment, and decide each line's visibility by checking whether it is hidden by material in front of it in that view.
The final sheet is submitted with construction lines erased or, conversely, with the whole construction left in heavy line, uneven line weights, and a layout that runs off the format or crowds the title block.
The candidate works at full pressure with one pencil and one line weight, and leaves the layout decision to the end, so the composition and graphic hierarchy are never controlled.
What to do instead
Plan the layout with a light bounding box for each view before drawing, keep two pencils (hard for construction, soft for final) or two line weights, ink only the final outlines and axes, and leave construction lines as thin and faint but present where they show the method; check margins and title block before inking.