摘要
利用实际采集的系列试样 ,在工业固体核磁Magmonitor(IMR)上建立了标准曲线 ;在IMR上运行标样的核磁信号后 ,对FID信号与实验数值之间的匹配使用 32个误差分析因子进行分析 ,并在Cregcon软件下运行回归分析 ,将结果转化成仪器预测模型。结果表明 ,测试相对偏差在 10 %以内 ,数据重现性好 ,而粉料与共聚物的测试误差有待改进。将IMR测试结果与不同测试方法进行比较 。
In this paper, a measurement method for polyolefin quality parameter on Industrial solid nuclear magnetic resonance of Magmonitor has been established.Based on a series of practical samples,their standard curve and matching NMR signals are analysed.In analysing the relationship between FID signals and lab data of samples,32-error factors are adopted,by which regression of curves is run on the software CREGCON and transmitted into the prediction model.Adopting these models on homopolymer or copolymer of polypropylene,measurements of the pellets and powders of samples are made.Results show that the relative deviation of the data is within 10% while the measurement error is different with the form of the samples.Due to unstability of powder samples,it is suggested that the powder samples should be treated(e.g.,degass)before being measured.
出处
《石油化工》
CAS
CSCD
北大核心
2002年第1期48-52,共5页
Petrochemical Technology
基金
国家重点基础研究发展规划项目
北京化工研究院基金资助项目
关键词
工业固体核磁
标准曲线
预测模型
32因子误差分析
industry magnetic resonance of Magmonitor
standard curve
prediction model
32-error factor analysis