{"id":8513,"date":"2026-06-16T09:00:00","date_gmt":"2026-06-16T08:00:00","guid":{"rendered":"https:\/\/uplastech.com\/mold-trial-first-article\/"},"modified":"2026-11-02T02:26:36","modified_gmt":"2026-11-02T02:26:36","slug":"mold-trial-first-article","status":"publish","type":"post","link":"https:\/\/uplastech.com\/ja\/mold-trial-first-article\/","title":{"rendered":"Mould Trial and First Article: What to Check, and in What Order"},"content":{"rendered":"<style>\n.upl-blog{font-size:17px;line-height:1.78;color:#1f2937}\n.upl-blog p{margin:0 0 18px}\n.upl-blog h2{font-size:26px;line-height:1.3;color:#0b2b33;margin:46px 0 14px;font-weight:700;letter-spacing:-.015em}\n.upl-blog h3{font-size:19.5px;color:#0d3b47;margin:28px 0 10px;font-weight:700}\n.upl-blog ul,.upl-blog ol{margin:0 0 20px;padding-left:22px}\n.upl-blog li{margin:0 0 9px}\n.upl-blog figure{margin:0 0 34px}\n.upl-blog figure img{width:100%;height:auto;display:block;border-radius:10px}\n.upl-blog figcaption{font-size:13px;color:#6b7280;margin-top:10px;line-height:1.5}\n.upl-blog .toc{background:none;border:0;border-radius:0;padding:0;margin:0 0 34px}\n.upl-blog .toc .lwptoc{margin:0}\n.upl-blog .toc h4{margin:0 0 10px;font-size:13px;letter-spacing:.1em;text-transform:uppercase;color:#0d5563}\n.upl-blog .toc ol{margin:0;padding-left:18px;font-size:14.5px}\n.upl-blog .toc li{margin-bottom:6px}\n.upl-blog table{width:100%;border-collapse:collapse;font-size:14.5px;margin:0 0 10px}\n.upl-blog th{background:#0d3b47;color:#fff;text-align:left;padding:11px 13px;font-weight:600;font-size:13.5px;border:0}\n.upl-blog td{padding:11px 13px;border:0;border-bottom:1px solid #e8ebed;vertical-align:top}\n.upl-blog tbody tr:nth-child(even){background:#fafbfc}\n.upl-blog .tablecap{font-size:12.5px;color:#6b7280;margin:0 0 30px}\n.upl-blog .myth{background:#fff8ef;border:1px solid #f6ddbd;border-radius:8px;padding:16px 20px;margin:0 0 26px;font-size:15.5px}\n.upl-blog .myth b{color:#a85c10}\n.upl-blog .callout{background:#f2f8f9;border-radius:8px;padding:18px 22px;margin:0 0 26px;font-size:15.5px;border-left:3px solid #14707f}\n.upl-blog .faq h3{border-top:1px solid #eceff1;padding-top:22px;margin-top:0;font-size:18px}\n.upl-blog .faq h3:first-of-type{border-top:0;padding-top:0}\n.upl-blog .cta{background:linear-gradient(135deg,#0b2b33,#0d5563);color:#fff;border-radius:12px;padding:34px;margin:48px 0 30px}\n.upl-blog .cta h2,.upl-blog .cta .cta-h{color:#fff;margin:0 0 10px;font-size:24px;font-weight:700;line-height:1.3}\n.upl-blog .cta p{color:#c9dee3;margin:0 0 20px;font-size:16px}\n.upl-blog .btn{display:inline-block;background:#e07a1f;color:#fff;text-decoration:none;font-weight:700;padding:13px 26px;border-radius:8px;font-size:15.5px}\n.upl-blog .rel{border-top:1px solid #e8ebed;margin-top:40px;padding-top:26px}\n.upl-blog .rel h4{font-size:13px;letter-spacing:.1em;text-transform:uppercase;color:#6b7280;margin:0 0 14px}\n.upl-blog .rel ul{list-style:none;padding:0;margin:0}\n.upl-blog .rel li{padding:11px 0;border-bottom:1px dashed #e8ebed}\n.upl-blog .rel a{text-decoration:none;font-weight:600;font-size:15.5px}\n@media(max-width:640px){.upl-blog h2{font-size:22px}.upl-blog table{font-size:13px}.upl-blog .cta{padding:24px}}\n\/* \u5355\u7bc7\u9875\u9690\u85cf WhatsApp \u95ee\u5019\u5361\uff0c\u53ea\u4fdd\u7559\u5706\u5f62\u6c14\u6ce1\uff0c\u907f\u514d\u906e\u6321\u4fa7\u680f\u8be2\u76d8\u63d0\u4ea4\u6309\u94ae\u3002\n   \u6ce8\u610f\uff1aClick to Chat \u7684\u70b9\u51fb\u903b\u8f91\u662f\u300c\u95ee\u5019\u5361\u5143\u7d20\u5b58\u5728\u5c31\u4e0d\u8df3\u8f6c\u300d\uff0c\u53ea\u9760 CSS \u9690\u85cf\u4f1a\u8ba9\u6c14\u6ce1\u53d8\u6210\u6b7b\u94fe\u63a5\u3002\n   Code Snippets #9\u300cUPL single post: WhatsApp bubble jumps straight to chat\u300d\u8d1f\u8d23\u628a\u8be5\u8282\u70b9\u4ece DOM \u79fb\u9664\uff0c\n   \u4e0b\u9762\u8fd9\u6761 CSS \u53ea\u662f\u79fb\u9664\u524d\u7684\u89c6\u89c9\u515c\u5e95\u3002\u5220 #9 \u524d\u5fc5\u987b\u5148\u5220\u8fd9\u6761\u89c4\u5219\u3002 *\/\nbody.single-post .ht_ctc_chat_greetings_box{display:none !important}\n<\/style>\n<div class=\"upl-blog\">\n<figure>\n<img fetchpriority=\"high\" decoding=\"async\" src=\"https:\/\/uplastech.com\/wp-content\/uploads\/2026\/11\/mold-trial-first-article.jpg\" alt=\"Six identical grey injection moulded brackets lined up on a steel bench beside a digital caliper and a printed inspection sheet, with the moulding machine behind\" width=\"1536\" height=\"909\"><figcaption>Six parts from the same shot. The first article is a decision about which of them to measure.<\/figcaption><\/figure>\n<div class=\"toc\"><\/div>\n<h2 id=\"balance\">The tool is not ready until it is warm<\/h2>\n<p>A mould starts cold. The first shot fills a cavity at 25 C, the fortieth fills one at 80 C, and the parts are different in every dimension that depends on shrinkage. Measuring the first shot and adjusting the tool to it is the single most expensive mistake in toolmaking, because the correction is made against a transient.<\/p>\n<p>The chart on the left shows the deviation from the settled dimension over the first 160 shots of a cold start. The first shot is 0.18 mm off, the tenth is 0.08, the fortieth is 0.03 and by 80 shots the part is within 0.02 mm of its final size.<\/p>\n<p>The practical rule is to run 30 to 40 shots with no adjustment, then check whether the last five are stable within the tolerance band before measuring anything. If the dimensions are still drifting, the mould temperature controller is undersized, the cooling circuits are unbalanced, or the machine is not holding its settings.<\/p>\n<p>That waiting period is also the time to check the process window. Short shots at low pressure, flash at high pressure, and the pressure at which the cavity is just filled define a window, and a part made in the middle of that window is more robust than one made at its edge.<\/p>\n<figure>\n<img decoding=\"async\" src=\"https:\/\/uplastech.com\/wp-content\/uploads\/2026\/11\/mold-trial-first-article-chart.jpg\" alt=\"Dimensional deviation falling from 0.18 mm on the first shot to 0.02 mm after 80 shots, and a bar chart of what a mould trial typically finds\" width=\"1536\" height=\"909\"><figcaption>Left: the part settles over the first forty shots. Right: what a trial actually turns up.<\/figcaption><\/figure>\n<h2 id=\"order\">The order of checks<\/h2>\n<p>Check the easy things first, because a trial is a sequence of eliminations and each check makes the next one cheaper. The order is: does it fill, does it eject, does it measure, does it look right, and does it work in the assembly. Doing them in the other order wastes time on cosmetic issues in a part that does not fit.<\/p>\n<p>Filling is first because a short shot makes every other measurement meaningless. Establish the pressure at which the cavity fills completely, then add 10 percent and treat that as the baseline for everything else.<\/p>\n<p>Ejection is second because a part that sticks or distorts on ejection cannot be measured. If the part is hard to eject at the trial, it will be worse in production when the operator is in a hurry, and the fix belongs in the tool rather than in the release agent.<\/p>\n<p>Dimensions come third, and only after thermal balance. Measure the features that matter to the assembly first, not the whole drawing: a part that fits is a usable part even if a cosmetic dimension is out.<\/p>\n<p>Appearance and function come last. A weld line or a flow mark is a defect to be recorded and addressed, but it does not stop the tool from being signed off for dimensional approval.<\/p>\n<h2 id=\"firstarticle\">What a first article should actually measure<\/h2>\n<p>A first article inspection is not a full dimensional report. It is a record of the dimensions that are critical to function, taken on parts from a stable process, with the process settings recorded alongside. A report with 80 dimensions and no settings cannot be used to decide anything.<\/p>\n<p>The critical dimensions are the ones that appear in the tolerance stack: the ones that set the fit to another part, the ones that set the wall thickness at the gate, and the ones that a customer will measure on receipt. Everything else can be checked on a sample.<\/p>\n<p>Take the parts from at least three shots, and from different cavities if the tool is multi cavity. A single part from a single cavity hides the cavity to cavity variation, which is often the largest single source of out of tolerance parts in production.<\/p>\n<p>Record the date, the machine, the material batch, the melt temperature, the mould temperature, the injection pressure, the hold pressure and the cooling time. Six months later, when a dimension drifts, that record is the only way to tell whether the tool has moved or the process has.<\/p>\n<table>\n<thead>\n<tr>\n<th>Stage<\/th>\n<th>What to check<\/th>\n<th>Decision it supports<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Fill<\/td>\n<td>Pressure at which the cavity just fills<\/td>\n<td>Baseline for the process window<\/td>\n<\/tr>\n<tr>\n<td>Ejection<\/td>\n<td>Sticking, drag marks, distortion<\/td>\n<td>Tool polish or draft change<\/td>\n<\/tr>\n<tr>\n<td>Thermal balance<\/td>\n<td>Last 5 of 40 shots within tolerance<\/td>\n<td>Whether to measure at all<\/td>\n<\/tr>\n<tr>\n<td>\u5bf8\u6cd5<\/td>\n<td>Stack up features, wall at the gate<\/td>\n<td>Tool correction or acceptance<\/td>\n<\/tr>\n<tr>\n<td>\u5916\u89b3<\/td>\n<td>Weld lines, flow marks, sinks<\/td>\n<td>Gate or cooling change<\/td>\n<\/tr>\n<tr>\n<td>\u7d44\u7acb<\/td>\n<td>Fit to the mating part<\/td>\n<td>Sign off or rework<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p class=\"tablecap\">Table 1 \u2014 The order of a trial, and what each stage decides<\/p>\n<h2 id=\"findings\">Tool fix or process fix<\/h2>\n<p>The most useful judgement at a trial is whether a finding can be fixed with a setting or needs steel removed. A short shot at the end of flow is usually a process fix. A short shot in the same place at any pressure is a tool fix, because the gate or the runner is too small.<\/p>\n<p>Flash is the mirror image. Flash that appears only at high pressure is a process fix, and flash that appears at any pressure is a parting line problem that has to be corrected in the tool. Trying to run around it with a lower pressure produces a short shot somewhere else.<\/p>\n<p>Warpage is the finding most often misdiagnosed. It is usually a cooling or packing issue rather than a tool geometry issue, and it is fixed by balancing the cooling circuits or extending the hold time. Removing steel to compensate for warpage is the last resort and usually makes the part thinner in the wrong place.<\/p>\n<p>The chart on the right puts dimensional problems at 28 percent and short shots at 20. Together they are nearly half of all findings, and both are the ones where the tool versus process decision matters most.<\/p>\n<h2 id=\"record\">The record that saves the next trial<\/h2>\n<p>Every trial should produce a document that a different engineer could use to reproduce the part. That means the settings, the material, the mould temperature, the measured dimensions and the photographs of the defects. Without it, the next trial starts from zero.<\/p>\n<p>Photograph the defects with a scale in the frame. A description of a flow mark is worth little; a photograph with a ruler beside it tells the next person whether it is 2 mm or 20, which changes the diagnosis entirely.<\/p>\n<p>Keep the first article parts, labelled with the shot number. When a problem appears in production three months later, a part from shot 40 with a known dimension is the reference that settles whether the tool has moved.<\/p>\n<table>\n<thead>\n<tr>\n<th>Record<\/th>\n<th>Why it matters<\/th>\n<th>How long to keep<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Process settings sheet<\/td>\n<td>Reproduce the part<\/td>\n<td>Life of the tool<\/td>\n<\/tr>\n<tr>\n<td>Material batch and certificate<\/td>\n<td>Trace a property change<\/td>\n<td>Life of the tool<\/td>\n<\/tr>\n<tr>\n<td>Measured critical dimensions<\/td>\n<td>Detect tool drift<\/td>\n<td>Life of the tool<\/td>\n<\/tr>\n<tr>\n<td>Photographs with a scale<\/td>\n<td>Compare a later defect<\/td>\n<td>Life of the tool<\/td>\n<\/tr>\n<tr>\n<td>Retained first article parts<\/td>\n<td>The physical reference<\/td>\n<td>Until the tool is retired<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p class=\"tablecap\">Table 2 \u2014 What to keep, and for how long<\/p>\n<p>The retained parts are the cheapest item on the list and the one most often thrown away in the tidy up after a trial. A labelled bag of six parts in a drawer is worth more than a hundred photographs when a dimension is disputed.<\/p>\n<h2 id=\"signoff\">Signing off without regret<\/h2>\n<p>Sign off on a process, not on a part. The part in your hand was made at a particular set of temperatures and pressures, and if those are not written down and reproducible, the sign off covers nothing.<\/p>\n<p>Run a short capability check before sign off: 30 consecutive parts, measured on the critical dimensions, with the mean and the spread compared with the tolerance. A process that is centred but has a spread as wide as the tolerance will produce rejects as soon as the material batch changes.<\/p>\n<p>Finally, agree in writing what is still open. A tool with a known cosmetic defect at the gate, accepted for the first order, will be a dispute on the second. Recording it at sign off is not pessimism, it is the only way to keep the acceptance decision attached to the facts.<\/p>\n<div class=\"faq\">\n<h3>How many shots before measuring a new mould?<\/h3>\n<p>Thirty to forty, then check that the last five are stable. A mould reaches thermal balance in about 40 cycles; measuring before that records a transient and leads to tool corrections that are wrong.<\/p>\n<h3>What is the difference between a mould trial and a first article inspection?<\/h3>\n<p>The trial establishes that the tool can make the part and finds the process window. The first article inspection records the dimensions of parts made in that window. The trial comes first, and the inspection is only meaningful after it.<\/p>\n<h3>Should I measure every dimension at the first article?<\/h3>\n<p>No. Measure the dimensions in the tolerance stack and the wall at the gate, on parts from at least three shots and different cavities. A full report with 80 dimensions and no process settings cannot be used to decide anything.<\/p>\n<h3>Why does my part warp only in production?<\/h3>\n<p>Usually because the trial ran a longer cycle than production. A shorter cooling time leaves the part warmer at ejection, and the warpage appears. The cooling time is part of the process and has to be recorded and held.<\/p>\n<h3>Is flash a tool problem or a process problem?<\/h3>\n<p>If it appears only at high pressure, it is a process problem. If it appears at any pressure, the parting line is not closing and it is a tool problem. Running a lower pressure to avoid it usually creates a short shot elsewhere.<\/p>\n<h3>How long should I keep the first article parts?<\/h3>\n<p>Until the tool is retired. They are the physical reference for every later dispute about whether the tool has moved, and a labelled bag of parts costs nothing to keep.<\/p>\n<\/div>\n<div class=\"cta\">\n<p class=\"cta-h\">Bringing a new tool into production?<\/p>\n<p>Send the part drawing and the tool specification. We will tell you what the trial should establish, in what order, and which findings are likely to need steel.<\/p>\n<p><a class=\"btn\" href=\"https:\/\/uplastech.com\/ja\/custom-injection-molding\/\">Ask about mould trials<\/a>\n<\/div>\n<div class=\"rel\">\n<h4>Continue reading<\/h4>\n<ul>\n<li><a href=\"https:\/\/uplastech.com\/ja\/mold-flow-analysis\/\">Mould Flow Analysis: What It Predicts and What It Cannot<\/a><\/li>\n<li><a href=\"https:\/\/uplastech.com\/ja\/tolerance-stack-up-analysis\/\">Tolerance Stack Up Analysis: Why the Parts Do Not Fit<\/a><\/li>\n<li><a href=\"https:\/\/uplastech.com\/ja\/short-shot-injection-molding\/\">Short Shot in Injection Moulding: When the Cavity Will Not Fill<\/a><\/li>\n<li><a href=\"https:\/\/uplastech.com\/ja\/mold-temperature-control\/\">Mould Temperature Control: The Setting That Decides Everything<\/a><\/li>\n<li><a href=\"https:\/\/uplastech.com\/ja\/mold-tool-life\/\">Mould Tool Life: How Many Shots Before the Steel Moves<\/a><\/li>\n<\/ul>\n<\/div>\n<\/div>","protected":false},"excerpt":{"rendered":"<p>The first shot out of a new tool is not the part, and neither is the tenth. A mould needs forty cycles before it is thermally stable, and a first article taken before that is a measurement of a transient state.<\/p>","protected":false},"author":1,"featured_media":8511,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[35],"tags":[],"class_list":["post-8513","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-injection-molding"],"meta_box":{"post-to-quiz_to":[]},"_links":{"self":[{"href":"https:\/\/uplastech.com\/ja\/wp-json\/wp\/v2\/posts\/8513","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/uplastech.com\/ja\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/uplastech.com\/ja\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/uplastech.com\/ja\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/uplastech.com\/ja\/wp-json\/wp\/v2\/comments?post=8513"}],"version-history":[{"count":1,"href":"https:\/\/uplastech.com\/ja\/wp-json\/wp\/v2\/posts\/8513\/revisions"}],"predecessor-version":[{"id":8842,"href":"https:\/\/uplastech.com\/ja\/wp-json\/wp\/v2\/posts\/8513\/revisions\/8842"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/uplastech.com\/ja\/wp-json\/wp\/v2\/media\/8511"}],"wp:attachment":[{"href":"https:\/\/uplastech.com\/ja\/wp-json\/wp\/v2\/media?parent=8513"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/uplastech.com\/ja\/wp-json\/wp\/v2\/categories?post=8513"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/uplastech.com\/ja\/wp-json\/wp\/v2\/tags?post=8513"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}