The latest articles on DEV Community by Jacky Chan (@jackychan2026). https://dev.to/jackychan2026 https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F4217636%2F97802499-12ef-4942-b451-da43ed10554a.png https://dev.to/jackychan2026 en Jacky Chan Fri, 30 Oct 2026 16:07:32 +0000 https://dev.to/jackychan2026/calculating-epoxy-volume-for-irregular-shapes-the-segment-method-418d https://dev.to/jackychan2026/calculating-epoxy-volume-for-irregular-shapes-the-segment-method-418d <p>Most volume calculators assume a rectangle. Real epoxy projects — a live-edge river table, a void in a burl slab, an irregular crack — are not rectangles, and a single length × width × depth estimate can be off by 20% or more. Here is the segment method I use, and the maths behind it.</p> <h2> Divide and measure </h2> <p>Split the irregular channel into <code>n</code> segments along its length. For each segment <code>i</code>, measure the width <code>w_i</code> at that station. With segment length <code>L_i</code> and a constant pour depth <code>d</code>:<br> </p> <div class="highlight js-code-highlight"> <pre class="highlight plaintext"><code>V_raw = d × Σ (L_i × w_i) </code></pre> </div> <p>With 5–8 stations, this trapezoidal-style approximation is typically within a few percent of the true area — far better than one average-width guess, which systematically underestimates bulges and overestimates necks in a way that does not cancel out.</p> <h2> Raw volume is not the order quantity </h2> <p>Three corrections turn geometry into a purchase plan:</p> <ol> <li> <strong>Waste factor</strong> — 5–10% for mixing and transfer losses: <code>V_waste = V_raw × 1.10</code> </li> <li> <strong>Seepage / seal coat</strong> — porous live edge absorbs a measurable first coat. Budget it as a separate thin-coat volume over the wood contact area, not as a percentage fudge.</li> <li> <strong>Layer constraints</strong> — deep-pour resins have a maximum lift thickness (often ~2 in). Total volume does not change, but batch count and Part A / Part B split per batch do: <code>batches = ceil(depth / max_lift)</code>, and each batch is mixed at the product ratio (1:1, 2:1, 3:1 by volume or weight — never assume they are interchangeable).</li> </ol> <p>Unit discipline matters too: 1 US gallon = 231 cubic inches = 3.785 litres, and ounces in a product listing are fluid ounces, not weight ounces. Resin density (~1.1–1.2 g/ml for most casting resins) is why "weigh it instead" advice needs the specific product's density, not a generic 1.0.</p> <h2> Reference implementation </h2> <p>If you want to sanity-check your spreadsheet against a worked implementation of exactly this model — segment mode, waste, seepage, seal coat, and A/B split — this <a href="https://epoxyplanner.com/epoxy-calculator/" rel="noopener noreferrer">epoxy resin calculator</a> exposes its assumptions rather than hiding them behind a single number, which makes it useful as a test oracle for your own code.</p> <h2> Edge cases worth testing </h2> <ul> <li>Depth varying along the channel (then segments need their own <code>d_i</code>)</li> <li>A river that forks (treat each branch as its own segment chain)</li> <li>Metric inputs with imperial product sizes (convert once, at the end, not per segment)</li> </ul> <p>How do you handle seepage in your own estimators — fixed allowance, or measured after the seal coat? Curious how other builders model it.</p> diy tutorial math makers