How to Read a Mold Flow Report: 12 Checkpoints Before Cutting Steel

A 16-cavity connector tool went to cutting on a mold flow report nobody in the buying chain could read. The fill plots were green, and the first T1 shot showed 9 percent fill-time imbalance, two air traps on the cosmetic face and a warpage prediction of 0.9 mm per 100 mm that no one had checked. The rework cost 11 days and a full runner ring, all of it readable before the first spark of EDM. This article is the audit: 12 checkpoints you can run in 40 minutes on a report someone else produced.

Molding simulation software does not approve tools, people do, and a report nobody interrogates is a liability transferred onto your program schedule.

The Snapshot

• Fill balance under 5 percent. On multi-cavity tools the gap between the fastest and slowest runner path should stay under 5 percent of fill time. 8 to 12 percent imbalance is the most common finding in unaudited reports and it moves cavity-by-cavity weight by 0.05 to 0.2 g.
• Clamp margin of 10 to 15 percent. The projected clamp force from the pressure plot must sit below 85 to 90 percent of the machine rating, and peak injection pressure below the resin ceiling of typically 80 to 150 MPa.
• Warpage beyond 0.5 mm per 100 mm is a design flag. On a flat face that number points to gate or cooling redesign, not to a press setting. Do not accept a clamp fix for a simulation-predicted shape defect.
• Cooling delta above 3 to 5 C means an undercooled circuit. A water-side inlet-to-outlet rise that large, or a cavity face spread above 5 C on a visible or optical surface, is visible in the report tables before any steel is drilled.

Report Hygiene: What to Check Before the Plots

Before reading a single contour plot, check that the analysis describes the part you are buying. Simulation errors are rarely numerical; they are administrative. The three most expensive ones are the wrong CAD revision, the wrong resin grade and a mesh too coarse to see the wall you are about to build.

Open the first pages and confirm four facts. First, the model revision and part mass match the latest drawing issue, not the issue sent at RFQ. Second, the material is the exact commercial grade with the supplier datasheet number on file, not a generic entry from the solver library. Third, the solver, module set and version are named, because fill results between engines differ by 5 to 10 percent on long thin parts. Fourth, the mesh carries at least 3 elements across the nominal wall thickness, so a 2.5 mm wall needs elements of roughly 0.8 mm or finer.

Then read the cooling model declaration. A cooling analysis can model circuits as simple pipe elements or as mapped geometry against the actual drilled path, and the difference in predicted part temperature can reach 8 to 12 C on a tool with curved circuits. If the report does not say which method was used, the thermal checkpoints are guesswork. If you want the analysis workflow itself, from meshing to result reading inside the solver, the process side is covered in our mold flow analysis guide; this article is the audit you run on the output.

Checkpoints 1 and 2: Gate Plan and Fill Balance

Checkpoint 1: gate location and fill pattern symmetry. Read the fill time animation, not the still image, and ask where the last spot to fill lands. If the last-fill point sits on a visible face, the air trap that forms there is a cosmetic defect you have designed in. If two flow fronts meet head-on around a core pin, a weld line forms there, and its strength is typically 50 to 70 percent of the bulk material. A good report justifies the gate position against three things: hidden last-fill, weld lines away from load and appearance, and balanced path lengths. If the report never states why the gate is there, that is a finding.

Checkpoint 2: fill time balance across cavities. On multi-cavity tools, table the fill time for every cavity. The target is under 5 percent spread between the fastest and the slowest runner path. A common pattern is 1.80 s fastest against 1.95 s slowest, an 8.3 percent imbalance that leaves the late cavities packing at lower pressure. The consequence at T1 is a weight spread of 0.05 to 0.2 g per shot and short shots that appear only in two or three cavities, exactly the failure mode documented in our short shot troubleshooting guide. The fix before steel is a runner diameter change, a tab cutter trim of 0.1 to 0.3 mm on the long paths, or a gate land adjustment. Running injection pressure up at the trial only moves the imbalance into the pack.

Checkpoints 3 to 5: Pressure, Air Traps and Shear

Checkpoint 3: pressure distribution and clamp force. Two numbers decide machine fit. The first is peak injection pressure against the resin's recommended ceiling, typically 80 to 150 MPa for engineering thermoplastics; a plan that needs 160 MPa to fill PC/ABS through 0.6 mm walls is a wall-thickness or gate problem, not a stronger machine problem. The second is average cavity pressure multiplied by projected area. Take the reported average pressure in the fill phase, multiply by the part projected area on the parting line, and require the resulting clamp force to sit 10 to 15 percent below the press rating. A 240 cm2 projected area at 35 MPa average pressure asks for roughly 84 t of clamp, so a 100 t machine is the honest floor.

Checkpoint 4: air traps and weld lines on the projected surface. The solver can project both onto the finished part surface instead of the mid-mesh, and that projection is the page most reports skip. Every air trap that lands on a visible face, or sits above a section thicker than 3 mm where compressed gas raises local melt temperature and produces diesel-effect burns, needs an explicit vent decision in the report: vent depth 0.02 to 0.05 mm on the parting line, a relief pocket, or a gate move. Weld lines get the same treatment: locate them, then check them against load paths and appearance. Where a weld must exist, the first fixes are an overflow well at the convergence point, a thicker local section, or sequential valve-gate timing; our weld line reduction guide carries the detail.

Checkpoint 5: shear rate and shear heating near the gate. Pull the maximum shear rate plot and read the value at the gate and in the first 5 to 10 mm of runner. Above roughly 10,000 1/s on polycarbonate or POM the skin degrades: visible gloss streaks, gel specks, and molecular weight loss that a tensile bar from the gate area will expose. The first fix is a larger gate section or a shorter gate land, which stretches the melt path and cuts local shear. If the report shows only a single viscosity number, ask which apparent viscosity the solve used at the actual shear rate.

Checkpoints 6 to 8: Temperature, Cooling and Warpage

Checkpoint 6: temperature uniformity at the cavity surface. Read the part temperature at end of fill or at ejection, and take the spread across the visible face rather than the global maximum. The working target is under 5 C variation for glossy or optical surfaces and under 8 to 10 C for structural covers. A spread above that means the pack stage is freezing the part unevenly, storing differential shrinkage that appears as warpage at checkpoint 8 and as cavity-to-cavity weight drift.

Checkpoint 7: cooling efficiency and the inlet-to-outlet delta. The circuit table should list flow rate and water-side temperature rise for every loop. A rise above 3 to 5 C on a single circuit means the water leaves the tool warm and the far end of that circuit is undercooled relative to the design intent. The first fixes are forcing turbulence above a Reynolds number of roughly 10,000, which for an 8 to 12 mm drilled channel usually means 4 to 8 L/min, splitting long circuits, or adding a baffle where a straight gun drill leaves a hot core. Series-connected loops that report one combined delta hide this problem; demand per-circuit numbers.

Checkpoint 8: shrinkage and warpage prediction and direction. Two readings matter: magnitude and direction. Magnitude: total warpage above 0.5 mm per 100 mm of span on a flat face is a redesign flag for gate or cooling, not a clamp or press adjustment, because a predicted shape defect at end of cooling will not be clamped away in production. Direction: compare the predicted shrinkage along flow against cross-flow. For unfilled polypropylene the two sit near 1.2 to 1.8 percent; for 30 percent glass-filled nylon they split hard, roughly 0.3 percent along flow against 0.5 to 0.7 percent across it, which is why the gate plan decides the dimensions before the steel does. When the number breaches, the countermeasures are the ones in our thin wall warpage guide: move the gate toward the geometric center, even the pack, and symmetrize cooling, in that order.

Checkpoints 9 and 10: Sink Marks and Ejection

Checkpoint 9: sink marks against rib and boss thickness. The sink plot is predictive, so read it at the back of ribs, bosses and any local thickening. The design rule it audits is geometric: rib core-out and nominal thickness at 50 to 60 percent of the main wall, and bosses cored so the mass behind the surface does not exceed the sink threshold. Ribs left at 80 to 100 percent of wall thickness that light up the sink predictor mean the report is forecasting a defect the CAD could have avoided in an hour. A sink depth of 0.05 to 0.10 mm on a painted or textured face is visible at glancing light, and no hold-pressure setting removes it without flashing the parting line.

Checkpoint 10: ejection direction and undercuts flagged in the model. Confirm the parting line, the ejection direction and every lift, slide or unscrewing feature match the drawing, not a simplified model that quietly deletes them. Then read the ejection force estimate and the draft audit: textured faces need roughly 1 degree of draft per 0.025 mm of texture depth, so a VDI 27 surface around 25 microns deep wants 4 to 5 degrees minimum. An ejection force that climbs toward the kN range per pin on a deep-draw core is a texture or draft conversation to have before the cavity is sunk, and drag marks found at T1 cost far more to fix in finished inserts at HRC 50 to 54.

Checkpoints 11 and 12: Fiber Orientation and Runner Sizing

Checkpoint 11: fiber orientation where strength matters. For any glass-filled part that carries load, request the orientation tensor or skin-core map, then compare fiber direction against the load path. Tensile strength along flow in a 30 percent glass-filled PA6 runs 50 to 80 percent higher than across flow in the same coupon, and a weld line made of crossing fronts sits at the bottom of that spread. If the gate plan puts the primary load path perpendicular to flow, or puts a weld in the bend radius, the simulation did its job and the design has not yet. The fix is gate repositioning or part-line reorientation, decided on the report, not at the capability study.

Checkpoint 12: runner and gate sizing summary, volume balance and freeze order. The last read is the feed system bill of materials. Cold runner volume relative to shot weight tells you the regrind and cycle story: a runner above 30 percent of shot weight is capital burned every cycle. Then check gate freeze time against cavity pack time. If the runner or gate freezes first, the cavity loses pressure support early and sink and weight variation follow; if the gate holds open far past the pack end, you are over-packing and flashing. The report should list freeze times per gate and total runner volume; a summary that omits them has skipped the arithmetic that sizes your sprue bushing, hot runner shot capacity and press cycle.

The 12-Checkpoint Matrix

The table below compresses the twelve reads into a bench card: what to open in the report, the number that should stop the release, and the first fix to put in front of the designer. Print it and work top to bottom; a report that survives all twelve rows still gets the three questions in the next section.

← swipe to scroll →

Table note: red-flag numbers are typical industry thresholds for unfilled and filled semi-crystalline and amorphous thermoplastics. Confirm each against the resin datasheet and the part function before holding a supplier to it.

Three Questions That Interrogate the Analyst

Checkpoints read the report as delivered. The three questions below test whether the report is fragile, because a simulation is one assumption set deep and the analyst knows it. Ask them in writing and keep the answers in the tool file next to the drawing revisions.

Question 1: what would change your gate plan? Every experienced analyst has a second gate candidate; forcing it into the open shows whether the chosen gate is a decision or a default. If the answer is "nothing," the analysis is shallow. A real gate trade always carries at least one alternative worth modeling.

Question 2: what is the biggest assumption risk in this run? The standard answers are mesh coarseness, a generic material entry instead of the qualified grade, or a cooling model built from pipe elements rather than the actual drilled path. The answer identifies the page of the report you should re-read first, and it prevents the 8 to 12 C thermal error that a mis-declared cooling model hides until T1.

Question 3: show me the result with the mold 10 C hotter. Sensitivity beats a single green run. A process window that inverts at +10 C mold temperature, or a warpage sign that flips at the low end of hold pressure, means the part is not robust and the press will find that window edge on the second shift. Ask for the same plot with the resin lot MFI at its datasheet extreme and you bracket the two variables that drift in production. The trial plan that checks these predictions belongs in a documented sequence like our mold trial guide, starting at T1.

The Sign-Off Checklist

Sign the report only when every line below is true and dated. The signature makes the simulation your responsibility instead of the analyst's, so the 12-checkpoint audit happens first.

• Model identity: CAD revision, drawing issue and part mass match the released model, and the material is the exact commercial grade with its datasheet attached.
• Mesh declaration: at least 3 elements across the nominal wall, solver and module versions named, cooling method stated as mapped or pipe elements.
• Fill reads: last-fill hidden from visible faces, cavity fill spread under 5 percent, and no trap over a 3 mm section without a vent decision.
• Machine fit: clamp force at 10 to 15 percent below press rating, peak injection pressure under the resin ceiling of 80 to 150 MPa.
• Thermal: cavity face spread under 5 C on cosmetic parts, per-circuit water delta under 3 to 5 C at 4 to 8 L/min.
• Shape and surface: warpage under 0.5 mm per 100 mm, sink under the visible threshold at ribs and bosses, draft above 1 degree per 0.025 mm of texture.
• Sensitivity: at least one hot-mold and one low-pressure variant reviewed, with the analyst's answers to the three questions on file.
• Handover: the signed report and its assumptions are attached to the tool file, next to the DFM sign-off, and stay there for the life of the tool.

Mold Flow Report FAQ

Q1. How long should a real audit of a mold flow report take?

Forty to sixty minutes on a single-cavity part, most of it on the fill animation, the cavity table and the projected trap and weld pages. A report that cannot survive that window is not finished; the missing pages cost less to request now than the rework costs after cutting.

Q2. The simulation says warpage is 0.8 mm per 100 mm. Can the press fix it?

No, not durably. A predicted shape defect of that size on a flat face comes from asymmetric cooling or unbalanced pack, and clamp or hold-pressure adjustments trade warpage for flash, sink or internal stress. The fix lives in the gate plan, the circuit layout or the rib geometry, which is why checkpoint 8 blocks release instead of deferring to trial.

Q3. Is one fill run enough to sign a tool off?

No. Sign after at least one sensitivity variant: the same model with the mold 10 C hotter, or the hold pressure at the bottom of its working band. A plan that survives only its nominal run has a window narrower than press drift between shifts.

Q4. What fill imbalance is acceptable on a 16-cavity automotive connector tool?

Under 5 percent between the fastest and slowest cavity, since a connector carries tight dimensional and weight demands, and 8 to 12 percent imbalance shows at T1 as short shots in two or three cavities. IATF 16949 programs should see the cavity table in the report, not a single average fill time.

Q5. Should the buyer run the simulation or the mold shop?

Whoever runs it, the reader must sit on the build side. A precision tool shop that holds mold tolerances to ±0.005 mm needs the same discipline applied to the feed system numbers, and a designer who cannot read a trap projection cannot defend a gate plan at the design review.

The Bottom Line

A mold flow report is a set of numbers with a story attached, and the story is only trusted after the numbers are checked. Twelve reads decide the release: hidden last-fill, under 5 percent cavity balance, clamp force 10 to 15 percent below press rating, traps and welds projected onto the real surface, shear under 10,000 1/s at the gate, a cavity face inside 5 C, circuit deltas inside 3 to 5 C, warpage under 0.5 mm per 100 mm, ribs cored to 60 percent of wall, ejection and draft matched to the texture, fibers aligned with the load path, and a feed system whose freeze order the summary actually shows. Add the three questions and the dated sign-off, and the simulation earns its place ahead of the steel.

DieStrike builds precision injection and stamping tooling to IATF 16949 with ±0.005 mm mold tolerance, ±0.002 mm on EDM and jig grinding, HRC 62 max heat treatment and 120+ machines. Tool lead times run 2 to 5 weeks with standard components shipping in 3 to 7 days, and every program starts with a DFM review that reads the flow report line by line before a single plate is ordered.