Mold flow analysis is a simulation of what happens inside the cavity before the cavity exists. It predicts how the melt front advances, where the pressure peaks, where the part will warp and where air will trap — all on a computer, all before any steel is cut. The output is not a guarantee of a perfect part; it is a list of defects you no longer have to discover during trial, and a set of changes that are free at this stage and expensive later.
Key takeaways
- Mold flow analysis answers questions that can only otherwise be answered by cutting steel and running a trial.
- It predicts fill, packing, cooling, warpage, weld line position and air traps — the defects that cause most first-trial failures.
- The real value is in the changes it justifies: gate position, wall thickness, runner and cooling layout, all decided on screen.
- Simulation is only as good as the inputs. A rough mesh and a guessed material give a confident answer that is wrong.
- A written DFM report built on a flow study is what a buyer should be asking for, not a screenshot of a fill pattern.
In this guide
- 1. What mold flow analysis actually simulates
- 2. The five answers it gives you
- 3. How the study runs, step by step
- 4. Defects it catches before trial
- 5. Where the analysis changes the tooling decision
- 6. What limits its accuracy
- 7. How to read a mold flow report
- 8. How we use it
1. What mold flow analysis actually simulates
The software divides the part geometry into a mesh and then solves the flow of a polymer melt through it. The material is described by a viscosity model, the machine by an injection profile, and the tool by a cooling layout and a set of boundaries. The solver then steps through time: melt front position, pressure, temperature, shear rate and shear stress at every point, followed by packing and cooling.
What that gives you is a physics-based opinion about a design that does not exist yet. It is not a substitute for a toolmaker’s judgement, and it is not an oracle. It is a way of testing five or six design options in an afternoon instead of five or six tool modifications over three months.
2. The five answers it gives you
- Will the part fill? Fill pressure, the pressure needed at the gate, and whether the machine you have can deliver it. Flow length to wall thickness ratio and the short-shot risk fall out of this.
- From where should the part be fed? Gate location comparison, the effect of adding a second gate, and the consequences for weld line position.
- How should it be cooled? Cooling channel layout, cycle time estimate, and the areas that will stay hot and shrink differently from the rest.
- How much will it warp? Differential shrinkage from uneven cooling and fibre orientation in filled resins.
- Where are the weld lines and air traps? Position of flow-front meetings and the spots where air will be trapped against the cavity wall.
3. How the study runs, step by step
- Geometry cleanup. The 3D model is simplified for the mesh — tiny fillets, engraving and geometry that has no bearing on flow are removed. Poor cleanup is the commonest source of nonsense results.
- Mesh and material. The part is meshed, and the actual resin grade is selected from the material database rather than approximated. Viscosity data for the real grade is what makes the pressure prediction meaningful.
- Gate and runner definition. Gate position and size, runner layout, and the machine’s injection profile.
- Cooling definition. Channel geometry, water temperature and flow rate.
- Fill analysis. Melt front, pressure, temperature, shear.
- Pack analysis. Holding pressure and time, and how shrinkage is compensated across the part.
- Cooling and warpage analysis. Temperature distribution at ejection, and the resulting distortion.
- Iteration. Options are compared and the design is adjusted — then the whole cycle repeats on the improved version.
The last step is the one that pays. A study that stops after the first run has told you what is wrong; a study that iterates has fixed it.
4. Defects it catches before trial
| Defect | What the analysis shows | Typical fix |
|---|---|---|
| Short shot | Pressure required exceeds machine capability; frozen flow front | Move or add a gate, thicken the wall, hotter mould, higher-flow resin |
| Weld lines | Exactly where two flow fronts meet, and the angle between them | Reposition gates so the meeting point falls in a non-cosmetic or low-stress area |
| Air traps | Cells where air cannot escape before the front closes | Add vents or a flow leader, change gate position |
| Sink marks | Regions that stay hot and shrink after ejection | Local thickening adjustments, rib redesign, better cooling |
| Warpage | Distortion from uneven cooling or fibre orientation | Balance the cooling circuit, adjust gate position, review wall uniformity |
| Excessive shear | Shear rate and shear stress above material limits | Enlarge the gate, change the fill rate, resposition the gate |
| Long cycle | Which regions delay ejection temperature | Baffles, bubblers or conformal cooling in the hot spots |
5. Where the analysis changes the tooling decision
Mold flow is usually sold as a defect-prevention tool, which understates it. The more valuable use is at the point where it changes what you decide to build:
- Cavity count. A balanced 8-cavity tool needs a runner layout that fills every cavity simultaneously. Simulation is how you find out whether a proposed layout balances or whether you are about to build a tool that shorts on cavity six.
- Wall thickness. If a part cannot be filled at the designed wall, the answer may be a design change rather than a process change — and that decision belongs before tooling, not after.
- Runner versus hot runner. The pressure and material savings of a hot runner system can be quantified, which turns a preference into a number.
- Bridge tool versus production tool. If the analysis shows the design fills comfortably in a simple tool, an aluminium bridge tool becomes a low-risk purchase.
- Feasibility. Occasionally the analysis says the part cannot be moulded as drawn at all. That is the cheapest possible time to find out.
6. What limits its accuracy
Simulation is a model, and models fail in predictable ways:
- Wrong material data. Using a generic “ABS” instead of the actual grade invalidates pressure results. Viscosity curves differ, sometimes by a lot.
- Bad mesh. Too coarse, or over-simplified geometry, distorts flow paths and hides thin features.
- Rough cooling model. If the cooling circuit is approximated rather than modelled properly, the cycle time and warpage predictions are indicative only.
- Glass-filled resins. Fibre orientation prediction is the hardest part of the physics, and warpage results for filled materials should still be treated with caution.
- Tool reality. Simulation cannot model a tool with poor venting, a worn check ring or a mis-set temperature controller. Some “simulation was wrong” cases are actually process or maintenance problems.
The correct framing is that a good study removes most of the first-trial risk, not all of it. Trial is still needed to confirm.
7. How to read a mold flow report
Ask for four things and you can check the study’s quality yourself:
- The material used — named by grade, with a note if a substitute was assumed.
- Fill pressure versus machine capability — a percentage, not a picture. If the required pressure is at 95% of machine capacity, you are one variation away from a short shot.
- The gate options compared — a study that presents one gate position has not evaluated anything.
- The resulting design changes — what was changed in the part or the tool as a result. A flow study with no design consequence was decoration.
8. How we use it
Our answer: design work here is done in UG and checked in Moldflow, and the study is not sold as a separate line item — it is part of how a tool gets designed. The output you receive is a written DFM report within three working days, at no cost, containing the fill, pack and cooling results, the gate and runner options we compared, and the specific changes we recommend to the part or the tool. On a thin-wall or long-flow part, that is where gate position and ejection get settled; on a straightforward part it is shorter, and we will tell you when it is.
We build 20 to 30 sets a month in our own toolroom, machining critical mould dimensions to ±0.005 mm on 13 CNC machining centres up to 1,600 mm travel, six EDM machines including Sodick mirror-finish and twin-head units, and three wire EDMs. A tool cut to that accuracy holds ±0.01 mm on critical moulded dimensions, subject to the resin and the part geometry. Trial delivery is samples, trial photos, a trial video, the full process parameter sheet and a CMM dimensional report, traceable by mold number — which is also how we check the study against reality.
Related guides in this series
This article is part of our Mold Design & DFM Guide — a full walkthrough of the topic with the numbers and checklists behind each decision.
- Injection Molding Design Guide: 15 Essential Rules for Better Plastic Parts
- Design for Manufacturing: 7 Critical DFM Rules for Injection Molded Parts
- Injection Molding Draft Angle Guide: How Much Draft Do You Need?
- Injection Molding Defects FAQ: Causes, Prevention, and How to Fix Them
Get an engineering answer, not a sales pitch
If you are working on a mold flow analysis, send us the part and the target volume. We will come back with a design review, a tooling recommendation and an itemised quote — and if a different process or material is a better fit for your volume, we will say so. Talk to a manufacturing engineer.