The limitation of version 3.3 was that it was tied to a 15.556° C (60°F.) calibration temperature and that the density values were primarily under atmospheric conditions rather than in vacuum.
This regimen works well for US distilled spirits plants and laboratories verifying US blending operations. However, it did not meet the requirements of the international distilled spirits industry which use different calibration temperatures as well as different definitions of alcohol concentration as their standards. Nor did it meet the needs of laboratories which have hydrometers and instruments that are calibrated for in vacuum densities and who use scales or known masses in vacuum rather than scales in the atmosphere.
The original work undertaken by the Bureau of Standards was conducted on an in vacuum basis and the values and algorithms published by them fully conform with more modern investigations such as those undertaken by the International Organization of Metrology, a copy of whose published tables are located in the site Resources page.
It was only later that the Bureau of Standards altered its original work on behalf of the US Treasury to create the in Air tables that comprise the gauging manual. Since alcoholic proof is defined at 60° F. in Air even for international blenders, who sell to the US and other countries that use a proof standard need to translate their blending results into proof at 60° F. in Air.
The work of upgrading ABS has consisted of going back to the original research, confirming the accuracy of the values published compared to more modern investigations, deciphering the equations and programming the algorithms used to derive those results.
To confirm the results obtained and harmonize the air tables and the vacuum algorithms this version of ABS provides output values for both air density and in vacuum density. The user can blend either in air or vacuum and even adjust the atmospheric buoyancy to suit the air pressure prevailing at the time measurements are taken.
In order to keep things simple the original TTB functions of the program all remain as in Air values. A new section has been added for International Blending which contains the in Vacuum and/or Air functions.
How it is done: The algorithms developed by the original researchers made use of identifying a temperature invariant quality of an alcohol and water mixture, such as percent alcohol by mass, at a given temperature, and then using a least squares reduction of Alpha, Beta and Gamma values, invented by the researchers, to conform to the curves they developed to predict that quality at a second temperature. These formulas effect a reliable translation to any density from 4°C or absolute density to 40° and above.
In examining the original work it was also discovered that the researchers had published a very accurate volume correction algorithm that also used a least squares reduction and Alpha, Beta and Gamma values and was based on the knowing the percent alcohol by volume at a given temperature.
Since determining true proof at 15.556°C makes a very accurate determination of both percent alcohol by volume and percent by mass these quantities can be fed into the Bureau of Standards algorithms and used to predict in vacuum or in air densities at any other temperature from freezing to boiling point.
It is also possible to side step the true proof algorithm entirely and simply use absolute values as inputs and obtain in vacuum results as outputs. The benefit of using the algorithms is that the results are obtained without resort to any table lookup and so operate more continuously than any set of table data can provide.
In order to effectuate the universal application of these functions it has been necessary to conduct an investigation into the modern definitions relating to the density of water, the density of absolute alcohol, standard atmospheric pressure, as well as into the standard altitude, gravity and the physical constants used for conversion at the time the research was conducted with respect to the modern SI values currently in use.
It was also necessary to obtain a consistent and accurate set of conversion factors from one quantity to another and to apply the correct rounding in the correct sequence to obtain accurate results. The primary source for conversion values has been a software package known as Uconeer, published by Katmar Software but the Handbook of Chemistry and Physics, of various editions, and other sources have also been consulted and applied.
During the investigation it was discovered that there was a discrepancy in the air buoyancy values between the researchers and modern values. This discrepancy might have been caused by the composition of the counter weights used to balance the atmosphere. We can not know for sure but the air values determined might have been based upon steel weights being used rather than brass weights.
For instance the HCP 72nd edition defines the density of water determinations being conducted under the following circumstances.
The weights are for dry air at the same temperature as the water up to 40°C and at a barometric pressure corrected to 760 mm and against brass weights of 8.4 density at 0°C. Above 40°C the temperature of the air is assumed to be 20°C, i.e., the water is allowed to cool to 20°C prior to the weighings being made. The volumetric computations are based upon the relations that one liter = 1 dm3 and that 1 dm3 = 61.023744 in.3
Applying the specific gravity of brass against Steel
8.4 Brass
7.87 Steel
0.9369 Relative density of Brass to Iron.
It was found that the atmospheric buoyancy applied to the TTB tables was 93% of that used in modern definitions of atmospheric pressure.
0.0011510 Kg. per liter of Air HCP 72nd
93.00% What is % of Air Weight to use
0.0010704 Kg. per Liter of Air * % Factor = ABS Air Wt. per Liter
0.0000099 Dif.
9.9E-06 Difference as scientific notation
By knowing this is the case it is possible to add the correct amount of air to the in air densities to achieve the corresponding in vacuum density when reference to the tables is required.
The ability to know about and control air pressure has created a new feature of the program in the introduction of the ability to determine individual atmospheric pressure values as they affect density and to choose one’s own present or calibration air density when using instruments. It is hoped that a moist air density function will also be developed and implemented.