CRANIUM, Professional Edition v5 - Property Estimation Software

Version 5 of the Professional Edition of Cranium contains a new, integrated modeling document that enables you to create new estimation techniques from data, new regression techniques, and new exporters for ASPEN and MATLAB. See below for detailed descriptions of additions and updates.

Upgrade Procedure:

  • Installation: If you have a license for version 5.0 or 5.1, and would like to install or reinstall Cranium, you can follow the procedure for a minor upgrade. Click here for instructions.
  • Major Upgrade: If you are running Cranium 4.2 or earlier, you must perform a major upgrade. You will need to pay an upgrade fee, update your license key, and then download and install the new version. Click here for instructions.

If you are looking for an earlier version, click here.

Version 5.0, November 2023
New Capability: Cranium is now available in four editions

Cranium is now available in four editions:

  • Professional Edition, (formerly called the complete version) provides full capabilities for physical property management, property estimation, knowledge base creation, graph generation, report generation, and machine learning.
  • Basic Edition, (formerly called the reader or demonstration version) is freely available and provides most of the functionality of the Professional Edition. A key difference is that the Basic Edition cannot save any changes made to its documents.
  • WebServer Edition, enables Cranium's functionality to be accessed via a set of API requests. The WebServer Edition can be used to distribute physical property data and estimates via the internet.
  • Component Edition, enables Cranium's functionality to be accessed using a set of COM objects. The Component Edition can be used to provide physical property data and estimates to your own software application.
New Capability: A machine learning document is now available
Cranium can now create a new machine learning document that contains chapters for multiple linear regression and technique regression. The multiple linear regression chapter enables you to develop new group contribution techniques and regress new parameters for parameter-based techniques. The technique regression chapter enables you to extend current estimation techniques, e.g., by regressing a value for a new group's contribution or by fitting binary interaction parameters for an activity coefficient model.
Cranium's machine learning process begins by compiling physical property values, data and estimates, from a knowledge base. Cranium's programming language enables you to tailor this compilation based on property values, composition limits, molecular substructures, and many more characteristics. These compiled data are then regressed either to an existing estimation technique's model or a newly entered model. Finally, the regressed parameters are transferred back to Cranium's knowledge base for use in future estimations. Thus, Cranium has learned from this newly obtained knowledge.
New Capability: "Regression techniques" have been added

If data is available for a temperature or pressure dependent property, it is often possible to interpolate or extrapolate these values to obtain estimates at other temperatures or pressures. Cranium's new "Regression Techniques" perform such interpolations and extrapolations.

For example, the "Den,l(T): ZrOnTw Regression Technique", highlighted below, estimates the density at a new temperature by collecting data, regressing this data into a quadratic equation, examining the statistical significance of the regression, and then using the equation to calculate the density.

Those estimates generated by this technique are shown as green diamonds in the graph. Note that Cranium could not use this technique near the ends of the data range, due to insufficient available values, and thus selected another technique.

Version 5.1, March 2023
Update: MATLAB Exporter now uses MKS Classes
Cranium's MATLAB Exporter now generates properties files containing four MATLAB classes: MKSRequest; MKSEntity; MKSProperty; MKSDatum. These classes provide attributes and methods for storing and manipulating entity and property information as well as data and estimates. Using this object-oriented approach results in a properties file that is very simple to understand and very easy to use.
MKS MATLAB Properties File
Update: ASPEN Exporter now generates BKP files
Cranium's ASPEN Exporter now generates input files formatted as BKP files. Previously, Cranium generated INP formatted files, a much older format, that required extra steps within ASPEN to input the file.
ASPEN BKP Input File