If you have exact CHN analyses data you can relay on them and calculate  H to C ratio directly.

For sure you may /but do not have to/  burn the samples and from amounts of CO2 and H2O determined  you can calculate the H to C ratios.

The only condition for both methods is, the absolutely dry and impurities /incl. mineral ones/ free samples of soot to be used for analyses. Or, you have to know their exact contents in the samples to be able correct the resulting data.

I thnik H would convert to H2O so you probably can't .

Burn it by pure O2 and then measure the H2O/CO2 ratio

If you have exact CHN analyses data you can relay on them and calculate  H to C ratio directly.

For sure you may /but do not have to/  burn the samples and from amounts of CO2 and H2O determined  you can calculate the H to C ratios.

The only condition for both methods is, the absolutely dry and impurities /incl. mineral ones/ free samples of soot to be used for analyses. Or, you have to know their exact contents in the samples to be able correct the resulting data.

From a typical CHN(S) analysis, what seems you have done with your samples, the H and C values are exactly what you need fo H/C calculations. Just take care for weight and atomic ratio.

Generally chemical composition of diesel is assumed as C10H22, from which C/H ratio can be estimated. For preparation of biodiesel, vegetable oil is to be heated up to 50-60 deg C and is t o be reacted wih methanol in presence of  methanol. Some times two stage esterification is required depending on content of fatty acids. Vegetable oil composition is known to us. After esterfication, chemical composition can be calcualted The most important condtions to be maintained in esterification are reaction time, temperature and molal conc of methanol. Hence C/H value of biodiesel can be estimaed. Higher the value of C/H ratio, more CO2, more CO and hence more carbon particles in the exhaust of the engine.  

I would recommend using Soxhlet solvent extraction of an aliquot of the sample.  You can then conduct GC-FID or other GC analyses.  GC-FID can provide an alkane-equivalent carbon number distribution for the sample, but not hydrogen.  You may be able to reasonably infer C:H with additional exhaust sample analyses or characterization of the fuels and lubricants used.  Keep in mind that for engines with reasonably good combustion efficiency, the VOC present as PM is primarily exhaust products from  lubricant for the petroleum diesel fuel and a combination of exhaust products from both lubricant and fatty-acid methyl ester species in the case biodiesel fuel usage; assuming no use of catalytic exhaust aftertreatment.  Products of petroleum diesel fuel in the exhaust are primarily (though not entirely) in the gas phase at typical exhaust sampling temperatures (~50 C for air-diluted exhaust samples) and typical VOC concentrations in dilute exhaust.  A large fraction of the VOC emitted from non-catalyzed diesel exhaust is from lubricant and fuel trapped in crevice volumes and thus not fully exposed to combustion processes, so considerable information may be gained via fuel and lubricant analysis.

Two stage esterification-The first stage (acid-catalyzed) of the process is to reduce the free fatty acids (FFA) content in vegetable oil by esterification with methanol (99% pure) and acid catalyst (sulfuric acid-98% pure) in 1 hour time of reaction at 55°C. Molar ratio of vegetable oil to methanol was to be determined and 0.75% catalyst (w/w).  

In the second stage (alkali-catalyzed), the triglyceride portion of the vegetable oil will react with methanol and base catalyst (sodium hydroxide–99% pure), in one hour time of reaction at 65°C, to form methyl ester (biodiesel) and glycerol. To remove un–reacted methoxide present in raw methyl ester, it is purified by the process of water washing with air–bubbling.