-
E-mail
yingtai@sandegroup.com
-
Phone
15274996811
-
Address
101, San De Technology Production Building, 558 Tongzipo West Road, Changsha High tech Development Zone
Hunan Sande Yingtai Environmental Protection Technology Co., Ltd
Quality Control of Limestone Testing: Techniques for Using Standard Substances and Parallel Sample Measurement
Date: 2025-08-18Read: 0
Limestone testingQuality control is the core of ensuring data reliability, and the standardized use of reference materials and scientific determination of parallel samples are key means of controlling systematic and random errors, directly affecting the accuracy and repeatability of detection results.
The precise application of standard substances needs to be integrated throughout the entire testing process. The selection of standard substances that match the sample matrix is the primary principle: when detecting calcium carbonate content, calcium carbonate standard substances (such as GBW07203a) are preferred, and the deviation of their principal component content should be ≤ 0.5%; When determining impurity elements (iron, magnesium, silicon), a multi-element mixed standard solution should be used, with a concentration gradient covering 80% -120% of the expected value of the sample. The dilution of standard substances must strictly follow the step-by-step dilution method, for example, dilute a 1000 μ g/mL iron standard solution to 100 μ g/mL first, and then dilute it to 10 μ g/mL as needed, with the relative error of each dilution step controlled within ± 1%. The drawing of the standard curve should include at least 5 concentration points, with a correlation coefficient (R ²) ≥ 0.999, and should be validated with intermediate concentration points before daily testing. If the deviation exceeds 5%, the curve should be redrawn.
The determination and control of parallel samples are effective ways to evaluate the precision of detection. The preparation of parallel samples needs to be synchronized from the sample crushing stage to ensure that the particle size distribution (such as 95% passing through a 150 μ m sieve) and uniformity of the two samples are consistent. At least one set of parallel samples should be set for each batch of samples (≤ 20), and for key items (such as calcium carbonate content), it should be increased to two sets. During the measurement process, parallel samples need to be randomly inserted into the sample sequence to avoid human error. The relative deviation (RSD) of parallel samples should be set according to the project threshold: the RSD of principal component analysis (such as calcium carbonate) should be ≤ 2%, and the RSD of trace impurities (such as iron content<0.1%) can be relaxed to ≤ 10%. If the threshold is exceeded, the cause needs to be investigated from the aspects of sample preparation, reagent addition, instrument reading, etc. If necessary, resampling and determination should be carried out.
Collaborative verification of quality control can enhance data reliability. Combine standard substances with parallel samples, such as inserting one standard substance and one set of parallel samples in each batch of samples, and verify the accuracy of the method by checking whether the measured values of the standard substance are within the allowable range of the certificate; The accuracy of the batch data can only be determined to be valid if both are qualified through parallel sample deviation evaluation. For abnormal data, it is necessary to use spiked recovery experiments to further verify: adding a known amount of standard substance to the sample, the recovery rate should be between 95% -105%, to ensure that the detection method is free of matrix interference.
By standardizing the use of standard substances and the determination of parallel samples, it is possible toLimestone testingThe accuracy of the data has been improved to over 98%, providing reliable quality basis for its application in building materials, metallurgy and other fields, while laying a data foundation for production process optimization and product quality control.