# Gage Drift Time to Out-of-Tolerance Calculator > Source: https://netray.co/tools/gage-drift-time-to-out-of-tolerance-calculator > Publisher: Netray (netray.co), AI and integration engineering for ERP and manufacturing systems. > Section: Free Tools ## Short answer Project instrument error forward from observed drift to find months until the tolerance limit is reached and how much of the window a chosen interval uses. ## Category quality ## Industry Precision Manufacturing ## Audience - Metrologist - Calibration coordinator - Quality manager ## Tool **Kind:** calculator **Inputs:** #### Permissible error limit **Key:** limit **Type:** number **Default:** 0.01 **Min:** 0 **Step:** 0.001 **Unit:** mm **Help:** Maximum permissible error, as a positive value. #### As-left error at last calibration **Key:** err0 **Type:** number **Default:** 0.002 **Step:** 0.0005 **Unit:** mm **Help:** Use the absolute value of the error after adjustment. #### Observed drift rate **Key:** rate **Type:** number **Default:** 0.0006 **Step:** 0.0001 **Unit:** mm per month **Help:** Change in error per month from successive calibrations. #### Proposed calibration interval **Key:** interval **Type:** slider **Default:** 12 **Min:** 1 **Max:** 36 **Step:** 1 **Unit:** months #### Safety fraction of window **Key:** safety **Type:** select **Default:** 0.8 **Options:** ##### 70% **Value:** 0.7 ##### 80% **Value:** 0.8 ##### 90% **Value:** 0.9 **Outputs:** #### Months until limit reached **Key:** months_limit **Expr:** if(eq(rate, 0), 999, max(0, (limit - abs(err0)) / abs(rate))) **Format:** months **Description:** Shows 999 when no drift is observed. #### Projected error at end of interval **Key:** err_end **Expr:** abs(err0) + abs(rate) * interval **Format:** decimal #### Share of limit used at interval end **Key:** used **Expr:** err_end / limit * 100 **Format:** percent **Primary:** true **Description:** Over 100% means the gage is expected to be out of tolerance before it is recalibrated. #### Suggested interval **Key:** suggest **Expr:** floor(safety * months_limit) **Format:** months **Primary:** true **Description:** Safety fraction multiplied by months to limit. **Chart:** #### Share of error limit used vs interval **Type:** line **X:** interval **From:** 1 **To:** 36 **Steps:** 36 **Series:** - used **Disclaimer:** Linear drift is a simplification. Real intervals should follow your calibration system procedure and reliability history, for example ILAC-G24 or ISO 10012 methods. ## Content **Intro:** When two or three calibration records show the same instrument creeping in one direction, you can estimate when it will leave tolerance. This calculator projects error forward and converts it to a defensible interval. **Sections:** #### How to use it **Body:** Enter the permissible error limit for the instrument, the as-left error from the last calibration and the drift rate per month. Drift is best estimated from at least three calibrations: take the change in as-found error divided by the months between them. Choose a safety fraction so the interval ends before the predicted crossing. The result shows months to the limit, projected error at your proposed interval and a suggested interval. **Bullets:** - Use as-found data, not as-left, for the drift rate. - Use the absolute error so direction does not matter. - Revisit after each calibration as the data improves. #### The logic **Body:** The model is a straight line: error at time t equals the starting error plus drift rate times t. Months to the limit equals remaining margin divided by drift rate. A safety fraction of 80 percent means you recalibrate when 80 percent of the remaining margin has been used. It is the same idea as the staircase and control chart methods in calibration interval guidance, reduced to one calculation you can explain to an auditor. #### Worked example **Body:** A micrometer has a limit of 0.010 mm, was left at 0.002 mm and drifts 0.0006 mm per month. Margin is 0.008 mm, so it reaches the limit in about 13.3 months. A 12 month interval projects an error of 0.0092 mm, which is 92 percent of the limit, close for comfort. An 80 percent safety fraction suggests 10 months. If the drift were due to wear on a contact face, repair might be a better answer than a shorter interval. #### Common mistakes **Body:** Do not project from one data point. Drift between two calibrations may be adjustment noise or a one-off shock. Avoid ignoring uncertainty, the as-found error is itself uncertain. Do not lengthen intervals only because an instrument was found in tolerance once. Finally, drift is not always linear, some instruments settle quickly after repair and then stabilise, so use the model as a guide, not a law. **Cta:** If calibration history sits in separate files, we can help bring due dates, as-found results and interval decisions into your ERP maintenance or quality module. **Faq:** #### How many calibrations do I need to estimate drift? **Answer:** Three is a reasonable minimum, giving two intervals to compare. More is better. With only two points you cannot tell drift from noise or a one-off event. #### Should I use as-found or as-left error? **Answer:** Estimate drift from as-found error at each calibration, since it shows what happened over the interval. As-left is the starting point for the next interval. #### Does this replace a formal interval method? **Answer:** No. It is a simple screening estimate. Formal methods such as those in ILAC-G24 use reliability statistics across populations of similar instruments. #### What if drift is zero? **Answer:** The calculator shows 999 months because no limit crossing is predicted. In practice, keep a maximum interval set by risk, manufacturer advice or regulation. ## Search terms gage drift calculator, calibration drift rate, time to out of tolerance, calibration interval drift analysis, instrument drift per month, calibration history trend, out of tolerance prediction, calibration interval justification --- Talk to an engineer: https://netray.co/contact?intent=erp_ai | Machine catalog: https://netray.co/agent/catalog.json